library(groundhog)
## Loaded 'groundhog' (version:2.0.1) using R-4.2.1
## Tips and troubleshooting: https://groundhogR.com
## [1;36mgroundhog says:[0m
## [36m
##
##
## OUTDATED GROUNDHOG
## You are using version '2.0.1
## The current version is '2.1.0'
##
## You can read about the changes here: https://groundhogr.com/changelog
##
## Update by running:
## install.packages('groundhog')[0m
pkgs <- c("lmerTest", "ggeffects","r2glmm", "tidyverse","here", "sjPlot", "ggpubr", "wesanderson", "effectsize","broom.mixed","corrr","report", "ez", "ggdist")
groundhog.day <- '2022-07-25'
groundhog.library(pkgs, groundhog.day)
## Loading required package: lme4
## Loading required package: Matrix
##
## Attaching package: 'lmerTest'
## The following object is masked from 'package:lme4':
##
## lmer
## The following object is masked from 'package:stats':
##
## step
## [36mSuccesfully attached 'lmerTest_3.1-3'[0m
## [36mSuccesfully attached 'ggeffects_1.1.2'[0m
## [36mSuccesfully attached 'r2glmm_0.1.2'[0m
## Registered S3 methods overwritten by 'readr':
## method from
## as.data.frame.spec_tbl_df vroom
## as_tibble.spec_tbl_df vroom
## format.col_spec vroom
## print.col_spec vroom
## print.collector vroom
## print.date_names vroom
## print.locale vroom
## str.col_spec vroom
## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.3.6 ✔ purrr 0.3.4
## ✔ tibble 3.1.8 ✔ dplyr 1.0.9
## ✔ tidyr 1.2.0 ✔ stringr 1.4.0
## ✔ readr 2.1.2 ✔ forcats 0.5.1
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ tidyr::expand() masks Matrix::expand()
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ✖ tidyr::pack() masks Matrix::pack()
## ✖ tidyr::unpack() masks Matrix::unpack()
## [36mSuccesfully attached 'tidyverse_1.3.2'[0m
##
## here() starts at /Users/jacobelder/Documents/GitHub/SelfAnchoring/Study 2
##
## [36mSuccesfully attached 'here_1.0.1'[0m
##
## [36mSuccesfully attached 'sjPlot_2.8.10'[0m
##
## [36mSuccesfully attached 'ggpubr_0.4.0'[0m
##
## [36mSuccesfully attached 'wesanderson_0.3.6'[0m
##
## [36mSuccesfully attached 'effectsize_0.7.0'[0m
##
## [36mSuccesfully attached 'broom.mixed_0.2.9.4'[0m
##
## Registered S3 method overwritten by 'seriation':
## method from
## reorder.hclust gclus
##
## [36mSuccesfully attached 'corrr_0.4.3'[0m
##
## [36mSuccesfully attached 'report_0.5.1'[0m
##
## [36mSuccesfully attached 'ez_4.4-0'[0m
##
## [36mSuccesfully attached 'ggdist_3.2.0'[0m
here::i_am("Study 2/Analysis/SAs2_Analysis.Rmd")
## here() starts at /Users/jacobelder/Documents/GitHub/SelfAnchoring
#plotDir <- "/Volumes/Research Project/Trait_TestRetest/WeekTRT/plots/"
devtools::source_url("https://raw.githubusercontent.com/JacobElder/MiscellaneousR/master/corToOne.R")
## ℹ SHA-1 hash of file is "07e3c11d2838efe15b1a6baf5ba2694da3f28cb1"
devtools::source_url("https://raw.githubusercontent.com/JacobElder/MiscellaneousR/master/plotCommAxes.R")
## ℹ SHA-1 hash of file is "374a4de7fec345d21628a52c0ed0e4f2c389df8e"
devtools::source_url("https://raw.githubusercontent.com/JacobElder/MiscellaneousR/master/named.effects.ref.R")
## ℹ SHA-1 hash of file is "0a5b928a75d310573e96ab72631b77a8a8b9acb3"
# fullTest <- read.csv("../Cleaning/output/fullTest.csv")
# fullTrain <- read.csv("../Cleaning/output/fullTrain.csv")
# traitsFreqs <- read.csv("../Cleaning/output/traitFreqOverUnder.csv")
fullTest <- arrow::read_parquet("../Cleaning/output/fullTest.parquet")
fullTrain <- arrow::read_parquet("../Cleaning/output/fullTrain.parquet")
traitsFreqs <- arrow::read_parquet("../Cleaning/output/traitFreqOverUnder.parquet")
uSubs <- unique(fullTest$subID)
indDiffs <- fullTest[!duplicated(fullTest$subID),]
allPosCents <- read.csv("/Volumes/GoogleDrive/My Drive/Volumes/Research Project/Trait Network_Behaviral/generating network/output/allPosCents.csv")
fullTest$ingChoiceN <- as.factor(fullTest$ingChoiceN)
fullTest$novel <- as.factor(fullTest$novel)
fullTest$selfResp.Z <- scale(fullTest$selfResp)
fullTest$SE.Z <- scale(fullTest$SE)
fullTest$iSE.Z <- scale(fullTest$iSE)
fullTest$oSE.Z <- scale(fullTest$oSE)
fullTest$predicted.Z <- scale(fullTest$predicted)
fullTest$slope.Z <- scale(fullTest$slope)
fullTest$entropy.Z <- scale(fullTest$entropy)
fullTest$WSR.Z <- scale(fullTest$WSR)
fullTest$neighAveOutSE.Z <- scale(fullTest$neighAveOutSE)
fullTest$neighAveAllSE.Z <- scale(fullTest$neighAveAllSE)
fullTest$neighAveInSE.Z <- scale(fullTest$neighAveInSE)
fullTest$evalLOO.Z <- scale(fullTest$evalLOO)
fullTest$propCorrInLOO.Z <- scale(fullTest$propCorrInLOO)
fullTest$propCorrOutLOO.Z <- scale(fullTest$propCorrOutLOO)
#fullTest$propCorr.Z <- scale(fullTest$propCorr)
fullTest$desirability.Z <- scale(fullTest$desirability)
fullTest$er.Z <- scale(fullTest$er)
fullTest$inDegree.Z <- scale(fullTest$inDegree)
fullTest$outDegree.Z <- scale(fullTest$outDegree)
fullTest$Ent.Z <- scale(fullTest$Ent)
fullTest$outgroup <- as.factor(fullTest$outgroup)
fullTest$outgroup <- relevel(fullTest$outgroup,"Not UCR")
fullTest$OtherTherm <- rowMeans(fullTest[c("Therm_2","Therm_4")])
fullTest$InOtherTherm <- fullTest$Therm_1 - fullTest$OtherTherm
fullTest$OtherStatus <- rowMeans(fullTest[c("UCLA_Status","CSULA_Status")])
fullTest$InOtherStatus <- fullTest$UCR_Status - fullTest$OtherStatus
fullTest$InOutStatus <- ifelse(fullTest$outgroup == "UCLA", fullTest$InUCLAStatus,
ifelse(fullTest$outgroup == "CSU LA", fullTest$InCSULAStatus,
ifelse(fullTest$outgroup == "Not UCR", fullTest$InOtherStatus, NA)))
fullTest$InOutTherm <- ifelse(fullTest$outgroup == "UCLA", fullTest$InUCLATherm,
ifelse(fullTest$outgroup == "CSU LA", fullTest$InCSULATherm,
ifelse(fullTest$outgroup == "Not UCR", fullTest$InOtherTherm, NA)))
fullTest$novel <- as.factor(fullTest$novel)
levels(fullTest$novel) <- list("Trained" = "0", "Held Out" = "1")
fullTest$outgroup <- as.factor(fullTest$outgroup)
fullTest$outgroup <- relevel(fullTest$outgroup,"Not UCR")
PCA<- prcomp(na.omit(fullTest[c("predicted","neighAveOutSE")]),
center = TRUE,
scale. = TRUE)
fullTest$PCA[!is.na(fullTest$predicted) & !is.na(fullTest$neighAveOutSE)] <- PCA$x[,1]
propMatrix <- matrix(nrow=148,ncol=7)
for(i in 1:148){
traitDf <- subset(fullTest, Idx==i)
test <- t.test(as.numeric(traitDf$ingChoiceN)-1, mu=.50)
propMatrix[i, ] <- c(i, test$statistic, test$p.value, test$conf.int, test$estimate, test$parameter)
}
colnames(propMatrix) <- c("Idx", "stat", "p", "LCI", "UCI", "est", "param")
propMatrix <- as.data.frame(propMatrix)
propMatrix$trait <- traitsFreqs$trait[1:148]
propMatrix <- propMatrix[order(propMatrix$p),]
propMatrix
Are there differences in the perceptions of status between
universities?
UCR students perceive UCLA as being significantly higher status and
CSU LA students as being significantly lower status.
indDiffs$Status <- NULL
statusDf <- pivot_longer(indDiffs, cols=ends_with("_Status"), names_to="University", values_to="Status") %>% select(subID, University, Status) %>% drop_na()
statusDf$University <- gsub("_Status","", statusDf$University)
statusDf$University <- as.factor(statusDf$University)
statusDf$subID <- as.factor(statusDf$subID)
statusDf$University <- relevel(statusDf$University, ref = "UCR")
m <- lmer( scale(Status) ~ University + ( 1 | subID), data = statusDf)
summary(m)
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: scale(Status) ~ University + (1 | subID)
## Data: statusDf
##
## REML criterion at convergence: 1169.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.1951 -0.5157 0.0584 0.5731 2.8926
##
## Random effects:
## Groups Name Variance Std.Dev.
## subID (Intercept) 0.08838 0.2973
## Residual 0.33374 0.5777
## Number of obs: 599, groups: subID, 200
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) -0.14003 0.04594 548.12194 -3.048 0.00242 **
## UniversityCSULA -0.71545 0.05786 397.80645 -12.366 < 2e-16 ***
## UniversityUCLA 1.13170 0.05777 397.29760 19.590 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) UCSULA
## UnvrstCSULA -0.628
## UnvrstyUCLA -0.629 0.499
ggpredict(m, c("University")) %>% plot()

m<-ezANOVA(statusDf, dv = Status, wid = subID, between = as.factor(University) )
## Warning: The column supplied as the wid variable contains non-unique values
## across levels of the supplied between-Ss variables. Automatically fixing this by
## generating unique wid labels.
## Warning: Data is unbalanced (unequal N per group). Make sure you specified a
## well-considered value for the type argument to ezANOVA().
## Coefficient covariances computed by hccm()
m
## $ANOVA
## Effect DFn DFd F p p<.05 ges
## 1 University 2 596 410.2656 9.077901e-113 * 0.5792539
##
## $`Levene's Test for Homogeneity of Variance`
## DFn DFd SSn SSd F p p<.05
## 1 2 596 5.413247 569.5049 2.832544 0.05965564
ggplot(statusDf, aes(University, Status)) +
ggdist::stat_halfeye(adjust = .5, width = .7, .width = 0, justification = -.2, point_colour = NA) +
geom_boxplot(width = .2, outlier.shape = NA) +
geom_jitter(width = .05, alpha = .3) + labs(y="Perceived Status",x="University") + jtools::theme_apa()

Do people feel differ in how warmly they feel towards each
university?
UCR students feel the most warmth towards UCR, significantly less
towards UCLA, and the least towards CSU LA.
warmthDf <- pivot_longer(indDiffs, cols=starts_with("Therm_"), names_to="University", values_to="Warmth") %>% select(subID, University, Warmth) %>% drop_na()
warmthDf$University <- gsub("Therm_1","UCR", warmthDf$University)
warmthDf$University <- gsub("Therm_2","UCLA", warmthDf$University)
warmthDf$University <- gsub("Therm_4","CSU LA", warmthDf$University)
warmthDf$University <- as.factor(warmthDf$University)
warmthDf$subID <- as.factor(warmthDf$subID)
warmthDf$University <- relevel(warmthDf$University, ref = "UCR")
m <- lmer( scale(Warmth) ~ University + ( 1 | subID), data = warmthDf)
summary(m)
## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: scale(Warmth) ~ University + (1 | subID)
## Data: warmthDf
##
## REML criterion at convergence: 1331.7
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.1404 -0.5117 0.1239 0.6562 2.3007
##
## Random effects:
## Groups Name Variance Std.Dev.
## subID (Intercept) 0.08294 0.288
## Residual 0.64005 0.800
## Number of obs: 529, groups: subID, 197
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 0.51774 0.06120 514.28694 8.460 2.78e-16 ***
## UniversityCSU LA -1.30600 0.08721 364.74221 -14.976 < 2e-16 ***
## UniversityUCLA -0.41053 0.08272 344.19662 -4.963 1.09e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) UCSULA
## UnvrstCSULA -0.623
## UnvrstyUCLA -0.656 0.463
ggpredict(m, c("University")) %>% plot()

m<-ezANOVA(warmthDf, dv = Warmth, wid = subID, between = as.factor(University) )
## Warning: The column supplied as the wid variable contains non-unique values
## across levels of the supplied between-Ss variables. Automatically fixing this by
## generating unique wid labels.
## Warning: Data is unbalanced (unequal N per group). Make sure you specified a
## well-considered value for the type argument to ezANOVA().
## Coefficient covariances computed by hccm()
m
## $ANOVA
## Effect DFn DFd F p p<.05 ges
## 1 University 2 526 102.2002 3.182127e-38 * 0.279847
##
## $`Levene's Test for Homogeneity of Variance`
## DFn DFd SSn SSd F p p<.05
## 1 2 526 3624.271 90108.74 10.57815 3.132948e-05 *
indDiffs$groupHomoph.Z <- scale(indDiffs$groupHomoph)
ggplot(warmthDf, aes(University, Warmth)) +
ggdist::stat_halfeye(adjust = .5, width = .7, .width = 0, justification = -.2, point_colour = NA) +
geom_boxplot(width = .2, outlier.shape = NA) +
geom_jitter(width = .05, alpha = .3) + labs(y="Perceived Warmth",x="University") + jtools::theme_apa()

Correlates with Assortativity in Group Predictioons
Dialectical self and need to belong positively associated with group
assortativity. Social identification, self-esteem, identity-centrality,
public collectivr eself-est
indDiffs %>% select(groupHomoph, seHomoph, SStatus:IdImp) %>% corToOne(., "groupHomoph")
## [1] "All required packages attached"
##
## Correlation method: 'pearson'
## Missing treated using: 'pairwise.complete.obs'
## Note: Using an external vector in selections is ambiguous.
## ℹ Use `all_of(referenceVar)` instead of `referenceVar` to silence this message.
## ℹ See <https://tidyselect.r-lib.org/reference/faq-external-vector.html>.
## This message is displayed once per session.
indDiffs %>% select(groupHomoph, seHomoph, SStatus:IdImp) %>% plotCorToOne(., "groupHomoph")
## [1] "All required packages attached"
##
## Correlation method: 'pearson'
## Missing treated using: 'pairwise.complete.obs'

Differences in assortativity of group predictions
library(ez)
m<-ezANOVA(indDiffs[!is.na(indDiffs$groupHomoph),], dv=groupHomoph, wid=subID, between=outgroup)
## Warning: Converting "subID" to factor for ANOVA.
## Warning: Converting "outgroup" to factor for ANOVA.
## Warning: Data is unbalanced (unequal N per group). Make sure you specified a
## well-considered value for the type argument to ezANOVA().
## Coefficient covariances computed by hccm()
hsd <- TukeyHSD(aov(groupHomoph ~ outgroup, data=indDiffs[!is.na(indDiffs$groupHomoph),]))
print(m)
## $ANOVA
## Effect DFn DFd F p p<.05 ges
## 1 outgroup 2 194 18.22331 5.590415e-08 * 0.1581565
##
## $`Levene's Test for Homogeneity of Variance`
## DFn DFd SSn SSd F p p<.05
## 1 2 194 0.01989084 0.4295431 4.491777 0.01238983 *
print(hsd)
## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = groupHomoph ~ outgroup, data = indDiffs[!is.na(indDiffs$groupHomoph), ])
##
## $outgroup
## diff lwr upr p adj
## Not UCR-CSU LA -0.007553224 -0.03543940 0.02033295 0.7983771
## UCLA-CSU LA 0.057477795 0.02969826 0.08525733 0.0000064
## UCLA-Not UCR 0.065031019 0.03714485 0.09291719 0.0000003
ezPlot(indDiffs[!is.na(indDiffs$groupHomoph),], groupHomoph, wid=subID, between=outgroup, x=.(outgroup))
## Warning: Converting "subID" to factor for ANOVA.
## Warning: Converting "outgroup" to factor for ANOVA.
## Warning: Data is unbalanced (unequal N per group). Make sure you specified a
## well-considered value for the type argument to ezANOVA().
## Coefficient covariances computed by hccm()
## Warning in ezStats(data = data, dv = dv, wid = wid, within = within, within_full
## = within_full, : Unbalanced groups. Mean N will be used in computation of FLSD

indDiffs$groupHomoph.Z <- scale(indDiffs$groupHomoph)
m <- lm(groupHomoph.Z ~ outgroup, data=indDiffs)
summary(m)
##
## Call:
## lm(formula = groupHomoph.Z ~ outgroup, data = indDiffs)
##
## Residuals:
## Min 1Q Median 3Q Max
## -1.8578 -0.6006 -0.1222 0.4185 2.9193
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -0.2288 0.1135 -2.016 0.0452 *
## outgroupNot UCR -0.1031 0.1612 -0.640 0.5231
## outgroupUCLA 0.7845 0.1605 4.887 2.14e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9222 on 194 degrees of freedom
## (3 observations deleted due to missingness)
## Multiple R-squared: 0.1582, Adjusted R-squared: 0.1495
## F-statistic: 18.22 on 2 and 194 DF, p-value: 5.59e-08
ggpredict(m, c("outgroup")) %>% plot(show.title=F, add.data=T)

ggplot(indDiffs, aes(outgroup, groupHomoph.Z)) +
ggdist::stat_halfeye(adjust = .5, width = .7, .width = 0, justification = -.2, point_colour = NA) +
geom_boxplot(width = .2, outlier.shape = NA) +
geom_jitter(width = .05, alpha = .3)
## Warning: Removed 3 rows containing missing values (stat_slabinterval).
## Warning: Removed 3 rows containing non-finite values (stat_boxplot).
## Warning: Removed 3 rows containing missing values (geom_point).

(Attempt to) replicate prior work: Ingroup favoritism exhibited by
ascribing more positive traits to ingroup
Desirability predicts ingroup predictions.
m <- glmer( ingChoiceN ~ desirability.Z + ( desirability.Z | subID) + ( 1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
summary(m)
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: ingChoiceN ~ desirability.Z + (desirability.Z | subID) + (1 |
## trait)
## Data: fullTest
## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))
##
## AIC BIC logLik deviance df.resid
## 35470.1 35519.8 -17729.1 35458.1 29325
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -4.3124 -0.9143 0.4457 0.7630 2.8230
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## subID (Intercept) 0.76943 0.8772
## desirability.Z 0.05154 0.2270 0.58
## trait (Intercept) 0.14360 0.3789
## Number of obs: 29331, groups: subID, 200; trait, 148
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.54560 0.07082 7.704 1.32e-14 ***
## desirability.Z 0.26431 0.03751 7.047 1.83e-12 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## desirblty.Z 0.225
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.726 (1.502 – 1.983)
|
0.122
|
7.704
|
<0.001
|
|
desirability Z
|
1.303 (1.210 – 1.402)
|
0.049
|
7.047
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.77
|
|
τ00 trait
|
0.14
|
|
τ11 subID.desirability.Z
|
0.05
|
|
ρ01 subID
|
0.58
|
|
ICC
|
0.23
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.016 / 0.239
|
Desirability predicts ingroup predictions most strongly for not CSU
LA, less so for UCLA, and least for CSU LA.
m <- glmer( ingChoiceN ~ desirability.Z * outgroup + ( desirability.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
summary(m)
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: ingChoiceN ~ desirability.Z * outgroup + (desirability.Z | subID) +
## (outgroup | trait)
## Data: fullTest
## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))
##
## AIC BIC logLik deviance df.resid
## 34327.2 34451.5 -17148.6 34297.2 29316
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.3269 -0.8549 0.4170 0.7136 4.5129
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## subID (Intercept) 0.62040 0.7877
## desirability.Z 0.05315 0.2305 0.65
## trait (Intercept) 0.32728 0.5721
## outgroupCSU LA 0.43552 0.6599 -0.89
## outgroupUCLA 0.64980 0.8061 -0.12 -0.18
## Number of obs: 29331, groups: subID, 200; trait, 148
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.97510 0.11056 8.820 < 2e-16 ***
## desirability.Z 0.37058 0.06023 6.153 7.59e-10 ***
## outgroupCSU LA -0.23383 0.15084 -1.550 0.12109
## outgroupUCLA -1.04611 0.15608 -6.703 2.05e-11 ***
## desirability.Z:outgroupCSU LA -0.20015 0.07512 -2.664 0.00771 **
## desirability.Z:outgroupUCLA -0.07658 0.08493 -0.902 0.36722
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) dsrb.Z oCSULA otUCLA d.Z:CL
## desirblty.Z 0.286
## outgrpCSULA -0.736 -0.209
## outgropUCLA -0.602 -0.202 0.397
## dsr.Z:CSULA -0.228 -0.813 0.326 0.162
## dsrb.Z:UCLA -0.203 -0.352 0.148 0.278 0.117
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.651 (2.135 – 3.293)
|
0.293
|
8.820
|
<0.001
|
|
desirability Z
|
1.449 (1.287 – 1.630)
|
0.087
|
6.153
|
<0.001
|
|
outgroup [CSU LA]
|
0.791 (0.589 – 1.064)
|
0.119
|
-1.550
|
0.121
|
|
outgroup [UCLA]
|
0.351 (0.259 – 0.477)
|
0.055
|
-6.703
|
<0.001
|
desirability Z * outgroup [CSU LA]
|
0.819 (0.707 – 0.948)
|
0.061
|
-2.664
|
0.008
|
desirability Z * outgroup [UCLA]
|
0.926 (0.784 – 1.094)
|
0.079
|
-0.902
|
0.367
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.62
|
|
τ00 trait
|
0.33
|
|
τ11 subID.desirability.Z
|
0.05
|
|
τ11 trait.outgroupCSU LA
|
0.44
|
|
τ11 trait.outgroupUCLA
|
0.65
|
|
ρ01 subID
|
0.65
|
|
ρ01 trait.outgroupCSU LA
|
-0.89
|
|
ρ01 trait.outgroupUCLA
|
-0.12
|
|
ICC
|
0.25
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.061 / 0.296
|
ggpredict(m, c("desirability.Z","outgroup")) %>% plot(show.title=F) + xlab("Desirability") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="desirability.Z [all]"` to get smooth plots.

Replication of prior self-anchoring findings: Self-evaluations
predicting ingroup evaluations
No covariates
Self-descriptiveness predict ingroup predictions
m <- glmer( ingChoiceN ~ selfResp.Z + ( selfResp.Z | subID) + ( 1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.690 (1.480 – 1.931)
|
0.115
|
7.727
|
<0.001
|
|
selfResp Z
|
1.325 (1.236 – 1.419)
|
0.047
|
7.977
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.67
|
|
τ00 trait
|
0.14
|
|
τ11 subID.selfResp.Z
|
0.15
|
|
ρ01 subID
|
0.23
|
|
ICC
|
0.22
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
17399
|
|
Marginal R2 / Conditional R2
|
0.018 / 0.238
|
With covariates
m <- glmer( ingChoiceN ~ selfResp.Z + desirability.Z + ( selfResp.Z | subID) + ( 1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.693 (1.488 – 1.925)
|
0.111
|
8.001
|
<0.001
|
|
selfResp Z
|
1.316 (1.229 – 1.410)
|
0.046
|
7.830
|
<0.001
|
|
desirability Z
|
1.233 (1.161 – 1.308)
|
0.038
|
6.868
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.67
|
|
τ00 trait
|
0.09
|
|
τ11 subID.selfResp.Z
|
0.15
|
|
ρ01 subID
|
0.23
|
|
ICC
|
0.22
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
17399
|
|
Marginal R2 / Conditional R2
|
0.031 / 0.241
|
With desirability covariates and differences with outgroup
Self-evaluations are more predictive of ingroup choices for negation
and least predictive for CSU LA.
m <- glmer( ingChoiceN ~ selfResp.Z * outgroup + desirability.Z + ( selfResp.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
summary(m)
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: ingChoiceN ~ selfResp.Z * outgroup + desirability.Z + (selfResp.Z |
## subID) + (outgroup | trait)
## Data: fullTest
## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))
##
## AIC BIC logLik deviance df.resid
## 20515.1 20639.3 -10241.6 20483.1 17383
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.7795 -0.8366 0.4165 0.7034 4.4633
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## subID (Intercept) 0.5237 0.7237
## selfResp.Z 0.1320 0.3633 0.22
## trait (Intercept) 0.2724 0.5220
## outgroupCSU LA 0.4595 0.6778 -0.93
## outgroupUCLA 0.6077 0.7795 -0.12 -0.19
## Number of obs: 17399, groups: subID, 200; trait, 148
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.92908 0.10519 8.832 < 2e-16 ***
## selfResp.Z 0.39008 0.05952 6.554 5.62e-11 ***
## outgroupCSU LA -0.19067 0.14535 -1.312 0.1896
## outgroupUCLA -1.02031 0.14940 -6.829 8.53e-12 ***
## desirability.Z 0.17394 0.02367 7.348 2.02e-13 ***
## selfResp.Z:outgroupCSU LA -0.15741 0.08208 -1.918 0.0551 .
## selfResp.Z:outgroupUCLA -0.18701 0.08413 -2.223 0.0262 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) slfR.Z oCSULA otUCLA dsrb.Z sR.Z:L
## selfResp.Z 0.148
## outgrpCSULA -0.748 -0.107
## outgropUCLA -0.607 -0.104 0.393
## desirblty.Z 0.004 -0.018 0.010 -0.003
## slR.Z:CSULA -0.107 -0.720 0.150 0.075 -0.040
## slfR.Z:UCLA -0.105 -0.702 0.076 0.151 -0.004 0.502
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.532 (2.060 – 3.112)
|
0.266
|
8.832
|
<0.001
|
|
selfResp Z
|
1.477 (1.314 – 1.660)
|
0.088
|
6.554
|
<0.001
|
|
outgroup [CSU LA]
|
0.826 (0.622 – 1.099)
|
0.120
|
-1.312
|
0.190
|
|
outgroup [UCLA]
|
0.360 (0.269 – 0.483)
|
0.054
|
-6.829
|
<0.001
|
|
desirability Z
|
1.190 (1.136 – 1.246)
|
0.028
|
7.348
|
<0.001
|
selfResp Z * outgroup [CSU LA]
|
0.854 (0.727 – 1.003)
|
0.070
|
-1.918
|
0.055
|
selfResp Z * outgroup [UCLA]
|
0.829 (0.703 – 0.978)
|
0.070
|
-2.223
|
0.026
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.52
|
|
τ00 trait
|
0.27
|
|
τ11 subID.selfResp.Z
|
0.13
|
|
τ11 trait.outgroupCSU LA
|
0.46
|
|
τ11 trait.outgroupUCLA
|
0.61
|
|
ρ01 subID
|
0.22
|
|
ρ01 trait.outgroupCSU LA
|
-0.93
|
|
ρ01 trait.outgroupUCLA
|
-0.12
|
|
ICC
|
0.24
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
17399
|
|
Marginal R2 / Conditional R2
|
0.072 / 0.294
|
ggpredict(m, c("selfResp.Z", "outgroup")) %>% plot(show.title=F) + xlab("Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()

Is it merely just average self-descriptiveness of the trait?
Even while controlling for the average of all other participants’
evaluations on each trait, participants’ evaluations are still
predictive of ingroup predictions.
m <- glmer( ingChoiceN ~ selfResp.Z + evalLOO.Z + desirability.Z + ( evalLOO.Z + selfResp.Z + desirability.Z | subID) + ( 1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
summary(m)
## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: binomial ( logit )
## Formula: ingChoiceN ~ selfResp.Z + evalLOO.Z + desirability.Z + (evalLOO.Z +
## selfResp.Z + desirability.Z | subID) + (1 | trait)
## Data: fullTest
## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+05))
##
## AIC BIC logLik deviance df.resid
## 21030.4 21146.9 -10500.2 21000.4 17384
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -6.4521 -0.8920 0.4264 0.7479 3.1218
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## subID (Intercept) 0.70911 0.8421
## evalLOO.Z 0.02170 0.1473 -0.22
## selfResp.Z 0.13816 0.3717 0.24 -0.02
## desirability.Z 0.04506 0.2123 0.60 0.03 0.23
## trait (Intercept) 0.07885 0.2808
## Number of obs: 17399, groups: subID, 200; trait, 148
##
## Fixed effects:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.54011 0.06683 8.082 6.37e-16 ***
## selfResp.Z 0.25573 0.03466 7.378 1.61e-13 ***
## evalLOO.Z 0.15309 0.03428 4.465 7.99e-06 ***
## desirability.Z 0.17277 0.03518 4.911 9.08e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) slfR.Z eLOO.Z
## selfResp.Z 0.165
## evalLOO.Z -0.053 -0.114
## desirblty.Z 0.246 0.071 -0.333
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.716 (1.506 – 1.956)
|
0.115
|
8.082
|
<0.001
|
|
selfResp Z
|
1.291 (1.207 – 1.382)
|
0.045
|
7.378
|
<0.001
|
|
evalLOO Z
|
1.165 (1.090 – 1.246)
|
0.040
|
4.465
|
<0.001
|
|
desirability Z
|
1.189 (1.109 – 1.273)
|
0.042
|
4.911
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.71
|
|
τ00 trait
|
0.08
|
|
τ11 subID.evalLOO.Z
|
0.02
|
|
τ11 subID.selfResp.Z
|
0.14
|
|
τ11 subID.desirability.Z
|
0.05
|
|
ρ01 subID.evalLOO.Z
|
-0.22
|
|
ρ01 subID.selfResp.Z
|
0.24
|
|
ρ01 subID.desirability.Z
|
0.60
|
|
ICC
|
0.23
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
17399
|
|
Marginal R2 / Conditional R2
|
0.040 / 0.264
|
Does similarity-weighted self-evaluation average predict ingroup
choices?
m <- glmer( ingChoiceN ~ WSR.Z + ( WSR.Z | subID) + ( 1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.636 (1.342 – 1.994)
|
0.165
|
4.878
|
<0.001
|
|
WSR Z
|
1.871 (1.433 – 2.443)
|
0.255
|
4.607
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.29
|
|
τ00 trait
|
0.17
|
|
τ11 subID.WSR.Z
|
2.59
|
|
ρ01 subID
|
0.06
|
|
ICC
|
0.55
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.051 / 0.575
|
Covariates
m <- glmer( ingChoiceN ~ WSR.Z + desirability.Z + ( WSR.Z | subID) + ( 1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.639 (1.351 – 1.990)
|
0.162
|
4.999
|
<0.001
|
|
WSR Z
|
1.846 (1.416 – 2.407)
|
0.250
|
4.529
|
<0.001
|
|
desirability Z
|
1.249 (1.174 – 1.329)
|
0.039
|
7.059
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.28
|
|
τ00 trait
|
0.12
|
|
τ11 subID.WSR.Z
|
2.58
|
|
ρ01 subID
|
0.07
|
|
ICC
|
0.55
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.057 / 0.573
|
Covariates and Condition Differences
Similarity-wieghted self-evaluations predict later ingroup
endorsements most for those with UCLA comparison and least for those
with CSU LA comparison.
m <- glmer( ingChoiceN ~ WSR.Z * outgroup + desirability.Z + ( WSR.Z + desirability.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.154 (1.648 – 2.815)
|
0.294
|
5.618
|
<0.001
|
|
WSR Z
|
1.561 (1.125 – 2.166)
|
0.261
|
2.665
|
0.008
|
|
outgroup [CSU LA]
|
0.981 (0.691 – 1.394)
|
0.176
|
-0.106
|
0.915
|
|
outgroup [UCLA]
|
0.454 (0.316 – 0.651)
|
0.084
|
-4.294
|
<0.001
|
|
desirability Z
|
1.235 (1.163 – 1.312)
|
0.038
|
6.856
|
<0.001
|
|
WSR Z * outgroup [CSU LA]
|
1.114 (0.692 – 1.796)
|
0.271
|
0.445
|
0.656
|
|
WSR Z * outgroup [UCLA]
|
0.822 (0.511 – 1.323)
|
0.200
|
-0.806
|
0.420
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.66
|
|
τ00 trait
|
0.34
|
|
τ11 subID.WSR.Z
|
1.26
|
|
τ11 subID.desirability.Z
|
0.05
|
|
τ11 trait.outgroupCSU LA
|
0.46
|
|
τ11 trait.outgroupUCLA
|
0.65
|
|
ρ01
|
0.07
|
|
|
0.59
|
|
|
-0.89
|
|
|
-0.16
|
|
ICC
|
0.42
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.066 / 0.459
|
ggpredict(m, c("WSR.Z", "outgroup")) %>% plot(show.title=F) + xlab("Similarity-Weighted Self-Evaluations") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="WSR.Z [all]"` to get smooth plots.

# fullTest$outgroupP <- factor(fullTest$outgroup, levels = c("CSU LA", "Not UCR", "UCLA"))
# contrasts(fullTest$outgroupP) <- contr.poly(3)
# m <- glmer( ingChoiceN ~ WSR.Z * outgroupP + desirability.Z + evalLOO.Z + ( WSR.Z | subID) + ( outgroupP | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
# optCtrl = list(maxfun = 100000)),
# nAGQ = 1)
# tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
# ggpredict(m, c("WSR.Z", "outgroup")) %>% plot(show.title=F) + xlab("Similarity-Weighted Self-Evaluations") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
Do cross-validation predictions predict ingroup choices?
No covariates
m <- glmer( ingChoiceN ~ predicted.Z + ( predicted.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.704 (1.478 – 1.966)
|
0.124
|
7.329
|
<0.001
|
|
predicted Z
|
1.581 (1.324 – 1.889)
|
0.143
|
5.058
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.51
|
|
τ00 trait
|
0.42
|
|
τ11 subID.predicted.Z
|
0.56
|
|
τ11 trait.outgroupCSU LA
|
0.47
|
|
τ11 trait.outgroupUCLA
|
0.69
|
|
ρ01 subID
|
0.28
|
|
ρ01 trait.outgroupCSU LA
|
-0.88
|
|
ρ01 trait.outgroupUCLA
|
-0.14
|
|
ICC
|
0.31
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.042 / 0.341
|
With covariates
m <- glmer( ingChoiceN ~ predicted.Z + desirability.Z + ( predicted.Z | subID) + ( 1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.622 (1.410 – 1.866)
|
0.116
|
6.774
|
<0.001
|
|
predicted Z
|
1.545 (1.287 – 1.855)
|
0.144
|
4.661
|
<0.001
|
|
desirability Z
|
1.260 (1.182 – 1.344)
|
0.041
|
7.052
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.52
|
|
τ00 trait
|
0.13
|
|
τ11 subID.predicted.Z
|
0.64
|
|
ρ01 subID
|
0.28
|
|
ICC
|
0.28
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.051 / 0.320
|
Covariates and Condition Differences
No effect
m <- glmer( ingChoiceN ~ predicted.Z*outgroup + desirability.Z + ( predicted.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.357 (1.877 – 2.960)
|
0.274
|
7.382
|
<0.001
|
|
predicted Z
|
1.466 (1.164 – 1.846)
|
0.172
|
3.254
|
0.001
|
|
outgroup [CSU LA]
|
0.864 (0.641 – 1.164)
|
0.132
|
-0.962
|
0.336
|
|
outgroup [UCLA]
|
0.404 (0.296 – 0.551)
|
0.064
|
-5.720
|
<0.001
|
|
desirability Z
|
1.197 (1.142 – 1.255)
|
0.029
|
7.505
|
<0.001
|
predicted Z * outgroup [CSU LA]
|
1.136 (0.790 – 1.633)
|
0.210
|
0.689
|
0.491
|
predicted Z * outgroup [UCLA]
|
0.874 (0.610 – 1.252)
|
0.160
|
-0.735
|
0.462
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.43
|
|
τ00 trait
|
0.34
|
|
τ11 subID.predicted.Z
|
0.36
|
|
τ11 trait.outgroupCSU LA
|
0.46
|
|
τ11 trait.outgroupUCLA
|
0.65
|
|
ρ01 subID
|
0.26
|
|
ρ01 trait.outgroupCSU LA
|
-0.90
|
|
ρ01 trait.outgroupUCLA
|
-0.14
|
|
ICC
|
0.27
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.077 / 0.324
|
ggpredict(m, c("predicted.Z", "outgroup")) %>% plot(show.title=F) + xlab("Cross-Validated Expectations") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="predicted.Z [all]"` to get smooth plots.

Does dot-product expected ratings predict ingroup choices?
No covariates
m <- glmer( ingChoiceN ~ er.Z + ( er.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.565 (1.306 – 1.876)
|
0.145
|
4.848
|
<0.001
|
|
er Z
|
1.399 (1.110 – 1.763)
|
0.165
|
2.846
|
0.004
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.01
|
|
τ00 trait
|
0.18
|
|
τ11 subID.er.Z
|
1.95
|
|
ρ01 subID
|
-0.03
|
|
ICC
|
0.49
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.017 / 0.497
|
With covariates
m <- glmer( ingChoiceN ~ er.Z + desirability.Z + ( er.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.568 (1.314 – 1.871)
|
0.141
|
4.984
|
<0.001
|
|
er Z
|
1.390 (1.103 – 1.751)
|
0.164
|
2.794
|
0.005
|
|
desirability Z
|
1.264 (1.186 – 1.347)
|
0.041
|
7.223
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.01
|
|
τ00 trait
|
0.13
|
|
τ11 subID.er.Z
|
1.95
|
|
ρ01 subID
|
-0.03
|
|
ICC
|
0.48
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.026 / 0.497
|
Covariates and Condition Differences
Marginall difference in CSU LA compared to Not UCR such that effect
of expected ratings is slightly stronger on ingroup choices.
m <- glmer( ingChoiceN ~ er.Z*outgroup + desirability.Z + ( er.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.120 (1.608 – 2.795)
|
0.299
|
5.330
|
<0.001
|
|
er Z
|
1.369 (0.986 – 1.901)
|
0.229
|
1.874
|
0.061
|
|
outgroup [CSU LA]
|
0.949 (0.660 – 1.366)
|
0.176
|
-0.280
|
0.780
|
|
outgroup [UCLA]
|
0.422 (0.291 – 0.613)
|
0.080
|
-4.531
|
<0.001
|
|
desirability Z
|
1.206 (1.150 – 1.264)
|
0.029
|
7.755
|
<0.001
|
|
er Z * outgroup [CSU LA]
|
1.196 (0.746 – 1.919)
|
0.288
|
0.743
|
0.457
|
|
er Z * outgroup [UCLA]
|
0.788 (0.491 – 1.265)
|
0.190
|
-0.987
|
0.324
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.65
|
|
τ00 trait
|
0.34
|
|
τ11 subID.er.Z
|
1.27
|
|
τ11 trait.outgroupCSU LA
|
0.47
|
|
τ11 trait.outgroupUCLA
|
0.65
|
|
ρ01 subID
|
0.02
|
|
ρ01 trait.outgroupCSU LA
|
-0.89
|
|
ρ01 trait.outgroupUCLA
|
-0.14
|
|
ICC
|
0.42
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.054 / 0.447
|
ggpredict(m, c("er.Z", "outgroup")) %>% plot(show.title=F) + xlab("Dot-Product Expected Rating") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="er.Z [all]"` to get smooth plots.

Does the linear trend of self-descriptiveness predict ingroup
choices?
No covariates
m <- glmer( ingChoiceN ~ slope.Z + ( slope.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.629 (1.357 – 1.956)
|
0.152
|
5.239
|
<0.001
|
|
slope Z
|
1.517 (1.188 – 1.938)
|
0.189
|
3.338
|
0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.01
|
|
τ00 trait
|
0.18
|
|
τ11 subID.slope.Z
|
2.13
|
|
ρ01 subID
|
-0.07
|
|
ICC
|
0.50
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.026 / 0.515
|
With covariates
m <- glmer( ingChoiceN ~ slope.Z + desirability.Z + ( slope.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.632 (1.365 – 1.951)
|
0.149
|
5.382
|
<0.001
|
|
slope Z
|
1.506 (1.180 – 1.922)
|
0.187
|
3.286
|
0.001
|
|
desirability Z
|
1.262 (1.184 – 1.345)
|
0.041
|
7.174
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.00
|
|
τ00 trait
|
0.13
|
|
τ11 subID.slope.Z
|
2.13
|
|
ρ01 subID
|
-0.07
|
|
ICC
|
0.50
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.034 / 0.514
|
Covariates and Condition Differences
Effect of self-descriptiveness trend is greatest for UCLA and weakest
for CSU LA
m <- glmer( ingChoiceN ~ slope.Z*outgroup + desirability.Z + ( slope.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.230 (1.699 – 2.925)
|
0.309
|
5.788
|
<0.001
|
|
slope Z
|
1.509 (1.087 – 2.095)
|
0.252
|
2.458
|
0.014
|
|
outgroup [CSU LA]
|
0.918 (0.643 – 1.309)
|
0.166
|
-0.474
|
0.635
|
|
outgroup [UCLA]
|
0.408 (0.283 – 0.587)
|
0.076
|
-4.827
|
<0.001
|
|
desirability Z
|
1.205 (1.149 – 1.264)
|
0.029
|
7.729
|
<0.001
|
slope Z * outgroup [CSU LA]
|
1.021 (0.635 – 1.644)
|
0.248
|
0.087
|
0.931
|
|
slope Z * outgroup [UCLA]
|
0.782 (0.486 – 1.259)
|
0.190
|
-1.012
|
0.312
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.59
|
|
τ00 trait
|
0.34
|
|
τ11 subID.slope.Z
|
1.24
|
|
τ11 trait.outgroupCSU LA
|
0.46
|
|
τ11 trait.outgroupUCLA
|
0.65
|
|
ρ01 subID
|
-0.02
|
|
ρ01 trait.outgroupCSU LA
|
-0.89
|
|
ρ01 trait.outgroupUCLA
|
-0.15
|
|
ICC
|
0.41
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.061 / 0.443
|
ggpredict(m, c("slope.Z", "outgroup")) %>% plot(show.title=F) + xlab("Similarity-Weighted Self-Evaluations") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="slope.Z [all]"` to get smooth plots.

Do the neighboring self-evaluations predict ingroup choices?
All neighbors
No covariates
m <- glmer( ingChoiceN ~ neighAveAllSE.Z + ( neighAveAllSE.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.690 (1.473 – 1.939)
|
0.118
|
7.491
|
<0.001
|
|
neighAveAllSE Z
|
1.236 (1.115 – 1.369)
|
0.065
|
4.052
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.62
|
|
τ00 trait
|
0.18
|
|
τ11 subID.neighAveAllSE.Z
|
0.36
|
|
ρ01 subID
|
-0.07
|
|
ICC
|
0.26
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29329
|
|
Marginal R2 / Conditional R2
|
0.010 / 0.269
|
With covariates
m <- glmer( ingChoiceN ~ neighAveAllSE.Z + desirability.Z + ( neighAveAllSE.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.692 (1.483 – 1.930)
|
0.114
|
7.807
|
<0.001
|
|
neighAveAllSE Z
|
1.232 (1.112 – 1.365)
|
0.064
|
4.001
|
<0.001
|
|
desirability Z
|
1.265 (1.187 – 1.347)
|
0.041
|
7.281
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.62
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveAllSE.Z
|
0.36
|
|
ρ01 subID
|
-0.07
|
|
ICC
|
0.25
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29329
|
|
Marginal R2 / Conditional R2
|
0.023 / 0.269
|
Covariates and Condition Differences
Effect of neighboring self-evaluations on ingroup choices is
strongest for UCLA and weakest for CSU LA
m <- glmer( ingChoiceN ~ neighAveAllSE.Z*outgroup + desirability.Z + ( neighAveAllSE.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.463 (1.991 – 3.047)
|
0.267
|
8.310
|
<0.001
|
|
neighAveAllSE Z
|
1.225 (1.050 – 1.429)
|
0.096
|
2.575
|
0.010
|
|
outgroup [CSU LA]
|
0.814 (0.612 – 1.082)
|
0.118
|
-1.415
|
0.157
|
|
outgroup [UCLA]
|
0.395 (0.294 – 0.530)
|
0.059
|
-6.169
|
<0.001
|
|
desirability Z
|
1.210 (1.154 – 1.269)
|
0.029
|
7.899
|
<0.001
|
neighAveAllSE Z * outgroup [CSU LA]
|
1.019 (0.819 – 1.267)
|
0.113
|
0.168
|
0.867
|
neighAveAllSE Z * outgroup [UCLA]
|
0.893 (0.716 – 1.115)
|
0.101
|
-0.999
|
0.318
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.50
|
|
τ00 trait
|
0.33
|
|
τ11 subID.neighAveAllSE.Z
|
0.25
|
|
τ11 trait.outgroupCSU LA
|
0.45
|
|
τ11 trait.outgroupUCLA
|
0.65
|
|
ρ01 subID
|
-0.02
|
|
ρ01 trait.outgroupCSU LA
|
-0.89
|
|
ρ01 trait.outgroupUCLA
|
-0.16
|
|
ICC
|
0.26
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29329
|
|
Marginal R2 / Conditional R2
|
0.052 / 0.299
|
ggpredict(m, c("neighAveAllSE.Z", "outgroup")) %>% plot(show.title=F) + xlab("Neighbors' Average Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="neighAveAllSE.Z [all]"` to get smooth plots.

All neighbors
No covariates
m <- glmer( ingChoiceN ~ neighAveOutSE.Z + ( neighAveOutSE.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.710 (1.493 – 1.958)
|
0.118
|
7.768
|
<0.001
|
|
neighAveOutSE Z
|
1.197 (1.100 – 1.302)
|
0.051
|
4.174
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.63
|
|
τ00 trait
|
0.18
|
|
τ11 subID.neighAveOutSE.Z
|
0.24
|
|
ρ01 subID
|
-0.08
|
|
ICC
|
0.24
|
|
N subID
|
200
|
|
N trait
|
147
|
|
Observations
|
29060
|
|
Marginal R2 / Conditional R2
|
0.007 / 0.248
|
With covariates
m <- glmer( ingChoiceN ~ neighAveOutSE.Z + desirability.Z + ( neighAveOutSE.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.710 (1.502 – 1.947)
|
0.113
|
8.100
|
<0.001
|
|
neighAveOutSE Z
|
1.195 (1.098 – 1.300)
|
0.051
|
4.145
|
<0.001
|
|
desirability Z
|
1.265 (1.188 – 1.347)
|
0.041
|
7.330
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.63
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveOutSE.Z
|
0.24
|
|
ρ01 subID
|
-0.08
|
|
ICC
|
0.23
|
|
N subID
|
200
|
|
N trait
|
147
|
|
Observations
|
29060
|
|
Marginal R2 / Conditional R2
|
0.021 / 0.249
|
Covariates and Condition Differences
Effect of neighboring self-evaluations on ingroup choices is
strongest for UCLA and weakest for CSU LA
m <- glmer( ingChoiceN ~ neighAveOutSE.Z*outgroup + desirability.Z + ( neighAveOutSE.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.523 (2.048 – 3.109)
|
0.269
|
8.689
|
<0.001
|
|
neighAveOutSE Z
|
1.151 (1.013 – 1.307)
|
0.075
|
2.161
|
0.031
|
|
outgroup [CSU LA]
|
0.804 (0.606 – 1.066)
|
0.116
|
-1.516
|
0.130
|
|
outgroup [UCLA]
|
0.382 (0.285 – 0.512)
|
0.057
|
-6.459
|
<0.001
|
|
desirability Z
|
1.209 (1.154 – 1.268)
|
0.029
|
7.894
|
<0.001
|
neighAveOutSE Z * outgroup [CSU LA]
|
1.086 (0.909 – 1.298)
|
0.099
|
0.908
|
0.364
|
neighAveOutSE Z * outgroup [UCLA]
|
0.943 (0.786 – 1.130)
|
0.087
|
-0.637
|
0.524
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.51
|
|
τ00 trait
|
0.33
|
|
τ11 subID.neighAveOutSE.Z
|
0.16
|
|
τ11 trait.outgroupCSU LA
|
0.45
|
|
τ11 trait.outgroupUCLA
|
0.63
|
|
ρ01 subID
|
-0.05
|
|
ρ01 trait.outgroupCSU LA
|
-0.89
|
|
ρ01 trait.outgroupUCLA
|
-0.15
|
|
ICC
|
0.25
|
|
N subID
|
200
|
|
N trait
|
147
|
|
Observations
|
29060
|
|
Marginal R2 / Conditional R2
|
0.052 / 0.287
|
ggpredict(m, c("neighAveOutSE.Z", "outgroup")) %>% plot(show.title=F) + xlab("Outwards Neighbors' Average Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="neighAveOutSE.Z [all]"` to get smooth plots.

Indegree neighbors
No covariates
m <- glmer( ingChoiceN ~ neighAveInSE.Z + ( neighAveInSE.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.777 (1.549 – 2.038)
|
0.124
|
8.208
|
<0.001
|
|
neighAveInSE Z
|
1.148 (1.064 – 1.238)
|
0.044
|
3.575
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.64
|
|
τ00 trait
|
0.43
|
|
τ11 subID.neighAveInSE.Z
|
0.16
|
|
τ11 trait.outgroupCSU LA
|
0.46
|
|
τ11 trait.outgroupUCLA
|
0.70
|
|
ρ01 subID
|
0.15
|
|
ρ01 trait.outgroupCSU LA
|
-0.87
|
|
ρ01 trait.outgroupUCLA
|
-0.15
|
|
ICC
|
0.27
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29282
|
|
Marginal R2 / Conditional R2
|
0.004 / 0.276
|
With covariates
m <- glmer( ingChoiceN ~ neighAveInSE.Z + desirability.Z + ( neighAveInSE.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.699 (1.488 – 1.940)
|
0.115
|
7.839
|
<0.001
|
|
neighAveInSE Z
|
1.158 (1.068 – 1.254)
|
0.047
|
3.569
|
<0.001
|
|
desirability Z
|
1.264 (1.186 – 1.348)
|
0.041
|
7.176
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.65
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveInSE.Z
|
0.20
|
|
ρ01 subID
|
0.06
|
|
ICC
|
0.23
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29282
|
|
Marginal R2 / Conditional R2
|
0.019 / 0.246
|
Covariates and Condition Differences
Effect of neighboring self-evaluations on ingroup choices is
strongest for UCLA and weakest for CSU LA
m <- glmer( ingChoiceN ~ neighAveInSE.Z*outgroup + desirability.Z + ( neighAveInSE.Z | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.498 (2.018 – 3.093)
|
0.272
|
8.402
|
<0.001
|
|
neighAveInSE Z
|
1.152 (1.014 – 1.309)
|
0.075
|
2.167
|
0.030
|
|
outgroup [CSU LA]
|
0.820 (0.614 – 1.094)
|
0.121
|
-1.352
|
0.176
|
|
outgroup [UCLA]
|
0.384 (0.285 – 0.517)
|
0.058
|
-6.294
|
<0.001
|
|
desirability Z
|
1.206 (1.150 – 1.264)
|
0.029
|
7.759
|
<0.001
|
neighAveInSE Z * outgroup [CSU LA]
|
1.015 (0.849 – 1.214)
|
0.093
|
0.168
|
0.866
|
neighAveInSE Z * outgroup [UCLA]
|
0.941 (0.785 – 1.129)
|
0.087
|
-0.652
|
0.514
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.54
|
|
τ00 trait
|
0.34
|
|
τ11 subID.neighAveInSE.Z
|
0.16
|
|
τ11 trait.outgroupCSU LA
|
0.46
|
|
τ11 trait.outgroupUCLA
|
0.65
|
|
ρ01 subID
|
0.12
|
|
ρ01 trait.outgroupCSU LA
|
-0.90
|
|
ρ01 trait.outgroupUCLA
|
-0.16
|
|
ICC
|
0.25
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29282
|
|
Marginal R2 / Conditional R2
|
0.050 / 0.291
|
ggpredict(m, c("neighAveInSE.Z", "outgroup")) %>% plot(show.title=F) + xlab("Inwards Neighbors' Average Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="neighAveInSE.Z [all]"` to get smooth plots.

Do people self-anchor more for higher indegree traits?
Similarity-Weighted Self-Evaluations
m <- glmer( ingChoiceN ~ WSR.Z * inDegree.Z + desirability.Z + ( WSR.Z + inDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.664 (1.371 – 2.020)
|
0.164
|
5.153
|
<0.001
|
|
WSR Z
|
1.858 (1.423 – 2.426)
|
0.253
|
4.554
|
<0.001
|
|
inDegree Z
|
1.003 (0.941 – 1.069)
|
0.033
|
0.092
|
0.927
|
|
desirability Z
|
1.249 (1.173 – 1.330)
|
0.040
|
6.938
|
<0.001
|
|
WSR Z * inDegree Z
|
1.037 (1.006 – 1.068)
|
0.016
|
2.345
|
0.019
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.27
|
|
τ00 trait
|
0.12
|
|
τ11 subID.WSR.Z
|
2.61
|
|
τ11 subID.inDegree.Z
|
0.01
|
|
ρ01 subID.WSR.Z
|
0.07
|
|
ρ01 subID.inDegree.Z
|
0.18
|
|
ICC
|
0.55
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.058 / 0.575
|
ggpredict(m, c("WSR.Z", "inDegree.Z")) %>% plot(show.title=F) + xlab("Similarity-Weighted Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="WSR.Z [all]"` to get smooth plots.

Cross-Validated Expectations
m <- glmer( ingChoiceN ~ predicted.Z * inDegree.Z + desirability.Z + ( predicted.Z + inDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.649 (1.436 – 1.894)
|
0.116
|
7.087
|
<0.001
|
|
predicted Z
|
1.550 (1.292 – 1.860)
|
0.144
|
4.710
|
<0.001
|
|
inDegree Z
|
1.001 (0.936 – 1.071)
|
0.034
|
0.039
|
0.969
|
|
desirability Z
|
1.261 (1.181 – 1.347)
|
0.042
|
6.951
|
<0.001
|
|
predicted Z * inDegree Z
|
1.034 (1.004 – 1.064)
|
0.016
|
2.194
|
0.028
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.51
|
|
τ00 trait
|
0.13
|
|
τ11 subID.predicted.Z
|
0.62
|
|
τ11 subID.inDegree.Z
|
0.01
|
|
ρ01 subID.predicted.Z
|
0.28
|
|
ρ01 subID.inDegree.Z
|
0.33
|
|
ICC
|
0.28
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.053 / 0.316
|
Dot-Product Expected Ratings
m <- glmer( ingChoiceN ~ er.Z * inDegree.Z + desirability.Z + ( er.Z + inDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.585 (1.328 – 1.891)
|
0.143
|
5.108
|
<0.001
|
|
er Z
|
1.400 (1.111 – 1.765)
|
0.165
|
2.849
|
0.004
|
|
inDegree Z
|
1.009 (0.944 – 1.078)
|
0.034
|
0.261
|
0.794
|
|
desirability Z
|
1.263 (1.184 – 1.347)
|
0.042
|
7.066
|
<0.001
|
|
er Z * inDegree Z
|
1.036 (1.006 – 1.067)
|
0.016
|
2.355
|
0.019
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.00
|
|
τ00 trait
|
0.13
|
|
τ11 subID.er.Z
|
1.96
|
|
τ11 subID.inDegree.Z
|
0.01
|
|
ρ01 subID.er.Z
|
-0.02
|
|
ρ01 subID.inDegree.Z
|
0.16
|
|
ICC
|
0.48
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.027 / 0.498
|
Linear Trend of Self-Descriptiveness
m <- glmer( ingChoiceN ~ slope.Z * inDegree.Z + desirability.Z + ( slope.Z + inDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.653 (1.384 – 1.976)
|
0.150
|
5.532
|
<0.001
|
|
slope Z
|
1.514 (1.185 – 1.934)
|
0.189
|
3.318
|
0.001
|
|
inDegree Z
|
1.009 (0.944 – 1.078)
|
0.034
|
0.260
|
0.795
|
|
desirability Z
|
1.261 (1.182 – 1.345)
|
0.042
|
7.021
|
<0.001
|
|
slope Z * inDegree Z
|
1.038 (1.008 – 1.070)
|
0.016
|
2.451
|
0.014
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.00
|
|
τ00 trait
|
0.13
|
|
τ11 subID.slope.Z
|
2.14
|
|
τ11 subID.inDegree.Z
|
0.01
|
|
ρ01 subID.slope.Z
|
-0.06
|
|
ρ01 subID.inDegree.Z
|
0.22
|
|
ICC
|
0.50
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.035 / 0.515
|
All Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveAllSE.Z * inDegree.Z + desirability.Z + ( neighAveAllSE.Z + inDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.697 (1.487 – 1.936)
|
0.114
|
7.864
|
<0.001
|
|
neighAveAllSE Z
|
1.249 (1.127 – 1.384)
|
0.065
|
4.241
|
<0.001
|
|
inDegree Z
|
1.012 (0.948 – 1.081)
|
0.034
|
0.365
|
0.715
|
|
desirability Z
|
1.262 (1.184 – 1.346)
|
0.041
|
7.121
|
<0.001
|
neighAveAllSE Z * inDegree Z
|
1.043 (1.012 – 1.074)
|
0.016
|
2.718
|
0.007
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.62
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveAllSE.Z
|
0.35
|
|
τ11 subID.inDegree.Z
|
0.01
|
|
ρ01 subID.neighAveAllSE.Z
|
-0.07
|
|
ρ01 subID.inDegree.Z
|
0.35
|
|
ICC
|
0.25
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29329
|
|
Marginal R2 / Conditional R2
|
0.025 / 0.271
|
Outwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveOutSE.Z * inDegree.Z + desirability.Z + ( neighAveOutSE.Z + inDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.713 (1.504 – 1.950)
|
0.114
|
8.117
|
<0.001
|
|
neighAveOutSE Z
|
1.201 (1.104 – 1.306)
|
0.052
|
4.249
|
<0.001
|
|
inDegree Z
|
1.011 (0.947 – 1.080)
|
0.034
|
0.337
|
0.736
|
|
desirability Z
|
1.263 (1.185 – 1.347)
|
0.041
|
7.170
|
<0.001
|
neighAveOutSE Z * inDegree Z
|
1.029 (0.999 – 1.059)
|
0.015
|
1.891
|
0.059
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.63
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveOutSE.Z
|
0.24
|
|
τ11 subID.inDegree.Z
|
0.01
|
|
ρ01 subID.neighAveOutSE.Z
|
-0.08
|
|
ρ01 subID.inDegree.Z
|
0.32
|
|
ICC
|
0.23
|
|
N subID
|
200
|
|
N trait
|
147
|
|
Observations
|
29060
|
|
Marginal R2 / Conditional R2
|
0.022 / 0.251
|
Inwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveInSE.Z * inDegree.Z + desirability.Z + ( neighAveInSE.Z + inDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.703 (1.492 – 1.945)
|
0.115
|
7.879
|
<0.001
|
|
neighAveInSE Z
|
1.186 (1.093 – 1.287)
|
0.049
|
4.099
|
<0.001
|
|
inDegree Z
|
1.018 (0.953 – 1.088)
|
0.034
|
0.535
|
0.593
|
|
desirability Z
|
1.260 (1.181 – 1.344)
|
0.042
|
6.986
|
<0.001
|
neighAveInSE Z * inDegree Z
|
1.051 (1.019 – 1.084)
|
0.017
|
3.142
|
0.002
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.66
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveInSE.Z
|
0.20
|
|
τ11 subID.inDegree.Z
|
0.01
|
|
ρ01 subID.neighAveInSE.Z
|
0.06
|
|
ρ01 subID.inDegree.Z
|
0.34
|
|
ICC
|
0.23
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29282
|
|
Marginal R2 / Conditional R2
|
0.021 / 0.248
|
Do people self-anchor more for higher outdegree traits?
Similarity-Weighted Self-Evaluations
Similarity-weighted self-descriptiveness more predictive of ingroup
choices for higher outdegree traits
m <- glmer( ingChoiceN ~ WSR.Z * outDegree.Z + desirability.Z + ( WSR.Z + outDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.689 (1.400 – 2.036)
|
0.161
|
5.481
|
<0.001
|
|
WSR Z
|
1.818 (1.409 – 2.345)
|
0.236
|
4.595
|
<0.001
|
|
outDegree Z
|
1.097 (1.017 – 1.182)
|
0.042
|
2.405
|
0.016
|
|
desirability Z
|
1.215 (1.134 – 1.301)
|
0.043
|
5.528
|
<0.001
|
|
WSR Z * outDegree Z
|
1.076 (1.034 – 1.121)
|
0.022
|
3.567
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.19
|
|
τ00 trait
|
0.12
|
|
τ11 subID.WSR.Z
|
2.31
|
|
τ11 subID.outDegree.Z
|
0.04
|
|
ρ01 subID.WSR.Z
|
0.04
|
|
ρ01 subID.outDegree.Z
|
0.48
|
|
ICC
|
0.53
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.059 / 0.554
|
ggpredict(m, c("WSR.Z", "outDegree.Z")) %>% plot(show.title=F) + xlab("Similarity-Weighted Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="WSR.Z [all]"` to get smooth plots.

Cross-Validated Expectations
m <- glmer( ingChoiceN ~ predicted.Z * outDegree.Z + desirability.Z + ( predicted.Z + outDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.699 (1.485 – 1.943)
|
0.116
|
7.740
|
<0.001
|
|
predicted Z
|
1.501 (1.297 – 1.738)
|
0.112
|
5.432
|
<0.001
|
|
outDegree Z
|
1.103 (1.018 – 1.195)
|
0.045
|
2.399
|
0.016
|
|
desirability Z
|
1.231 (1.145 – 1.323)
|
0.045
|
5.645
|
<0.001
|
|
predicted Z * outDegree Z
|
1.086 (1.041 – 1.134)
|
0.024
|
3.771
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.59
|
|
τ00 trait
|
0.13
|
|
τ11 subID.predicted.Z
|
0.23
|
|
τ11 subID.outDegree.Z
|
0.05
|
|
ρ01 subID.predicted.Z
|
0.13
|
|
ρ01 subID.outDegree.Z
|
0.86
|
|
ICC
|
0.23
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.056 / 0.275
|
Dot-Product Expected Ratings
m <- glmer( ingChoiceN ~ er.Z * outDegree.Z + desirability.Z + ( er.Z + outDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.616 (1.362 – 1.917)
|
0.141
|
5.507
|
<0.001
|
|
er Z
|
1.389 (1.114 – 1.731)
|
0.156
|
2.925
|
0.003
|
|
outDegree Z
|
1.101 (1.019 – 1.190)
|
0.044
|
2.432
|
0.015
|
|
desirability Z
|
1.228 (1.144 – 1.318)
|
0.045
|
5.657
|
<0.001
|
|
er Z * outDegree Z
|
1.066 (1.024 – 1.109)
|
0.022
|
3.141
|
0.002
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.94
|
|
τ00 trait
|
0.13
|
|
τ11 subID.er.Z
|
1.72
|
|
τ11 subID.outDegree.Z
|
0.04
|
|
ρ01 subID.er.Z
|
-0.05
|
|
ρ01 subID.outDegree.Z
|
0.55
|
|
ICC
|
0.46
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.030 / 0.478
|
Linear Trend of Self-Descriptiveness
m <- glmer( ingChoiceN ~ slope.Z * outDegree.Z + desirability.Z + ( slope.Z + outDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.677 (1.412 – 1.993)
|
0.148
|
5.879
|
<0.001
|
|
slope Z
|
1.497 (1.186 – 1.890)
|
0.178
|
3.399
|
0.001
|
|
outDegree Z
|
1.102 (1.020 – 1.191)
|
0.044
|
2.453
|
0.014
|
|
desirability Z
|
1.225 (1.141 – 1.315)
|
0.044
|
5.615
|
<0.001
|
|
slope Z * outDegree Z
|
1.074 (1.031 – 1.118)
|
0.022
|
3.456
|
0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.94
|
|
τ00 trait
|
0.13
|
|
τ11 subID.slope.Z
|
1.88
|
|
τ11 subID.outDegree.Z
|
0.04
|
|
ρ01 subID.slope.Z
|
-0.10
|
|
ρ01 subID.outDegree.Z
|
0.56
|
|
ICC
|
0.48
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.038 / 0.495
|
All Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveAllSE.Z * outDegree.Z + desirability.Z + ( neighAveAllSE.Z + outDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.720 (1.506 – 1.965)
|
0.117
|
7.997
|
<0.001
|
|
neighAveAllSE Z
|
1.267 (1.147 – 1.401)
|
0.065
|
4.644
|
<0.001
|
|
outDegree Z
|
1.121 (1.038 – 1.212)
|
0.044
|
2.901
|
0.004
|
|
desirability Z
|
1.222 (1.139 – 1.311)
|
0.044
|
5.614
|
<0.001
|
neighAveAllSE Z * outDegree Z
|
1.073 (1.033 – 1.115)
|
0.021
|
3.598
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.65
|
|
τ00 trait
|
0.12
|
|
τ11 subID.neighAveAllSE.Z
|
0.31
|
|
τ11 subID.outDegree.Z
|
0.05
|
|
ρ01 subID.neighAveAllSE.Z
|
-0.05
|
|
ρ01 subID.outDegree.Z
|
0.76
|
|
ICC
|
0.26
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29329
|
|
Marginal R2 / Conditional R2
|
0.030 / 0.279
|
Outwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveOutSE.Z * outDegree.Z + desirability.Z + ( neighAveOutSE.Z + outDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.738 (1.525 – 1.982)
|
0.116
|
8.273
|
<0.001
|
|
neighAveOutSE Z
|
1.234 (1.133 – 1.343)
|
0.054
|
4.845
|
<0.001
|
|
outDegree Z
|
1.109 (1.025 – 1.199)
|
0.044
|
2.590
|
0.010
|
|
desirability Z
|
1.228 (1.145 – 1.317)
|
0.044
|
5.769
|
<0.001
|
neighAveOutSE Z * outDegree Z
|
1.067 (1.026 – 1.110)
|
0.021
|
3.221
|
0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.65
|
|
τ00 trait
|
0.12
|
|
τ11 subID.neighAveOutSE.Z
|
0.22
|
|
τ11 subID.outDegree.Z
|
0.05
|
|
ρ01 subID.neighAveOutSE.Z
|
-0.09
|
|
ρ01 subID.outDegree.Z
|
0.78
|
|
ICC
|
0.24
|
|
N subID
|
200
|
|
N trait
|
147
|
|
Observations
|
29060
|
|
Marginal R2 / Conditional R2
|
0.028 / 0.262
|
Inwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveInSE.Z * outDegree.Z + desirability.Z + ( neighAveInSE.Z + outDegree.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.731 (1.512 – 1.981)
|
0.119
|
7.969
|
<0.001
|
|
neighAveInSE Z
|
1.169 (1.081 – 1.265)
|
0.047
|
3.895
|
<0.001
|
|
outDegree Z
|
1.120 (1.035 – 1.212)
|
0.045
|
2.802
|
0.005
|
|
desirability Z
|
1.224 (1.140 – 1.314)
|
0.044
|
5.557
|
<0.001
|
neighAveInSE Z * outDegree Z
|
1.053 (1.016 – 1.091)
|
0.019
|
2.857
|
0.004
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.69
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveInSE.Z
|
0.18
|
|
τ11 subID.outDegree.Z
|
0.05
|
|
ρ01 subID.neighAveInSE.Z
|
0.06
|
|
ρ01 subID.outDegree.Z
|
0.81
|
|
ICC
|
0.24
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29282
|
|
Marginal R2 / Conditional R2
|
0.024 / 0.262
|
Do cross-validated similarity * self-evaluation predictions predict
ingroup choices, regardless of whether it was seen prior or not?
Similarity-Weighted Self-Evaluations
m <- glmer( ingChoiceN ~ WSR.Z * novel + desirability.Z + ( WSR.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.651 (1.360 – 2.004)
|
0.163
|
5.065
|
<0.001
|
|
WSR Z
|
1.874 (1.436 – 2.446)
|
0.255
|
4.627
|
<0.001
|
|
novel [Held Out]
|
0.983 (0.932 – 1.036)
|
0.027
|
-0.643
|
0.520
|
|
desirability Z
|
1.249 (1.174 – 1.329)
|
0.039
|
7.064
|
<0.001
|
|
WSR Z * novel [Held Out]
|
0.965 (0.913 – 1.019)
|
0.027
|
-1.287
|
0.198
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.26
|
|
τ00 trait
|
0.12
|
|
τ11 subID.WSR.Z
|
2.59
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.WSR.Z
|
0.06
|
|
ρ01 subID.novelHeld Out
|
0.80
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.117 / NA
|
ggpredict(m, c("WSR.Z", "novel")) %>% plot(show.title=F) + xlab("Similarity-Weighted Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="WSR.Z [all]"` to get smooth plots.

Moderated by outgroup comparison
m <- glmer( ingChoiceN ~ WSR.Z * novel * outgroup + desirability.Z + ( WSR.Z + novel | subID) + ( outgroup | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.199 (1.637 – 2.953)
|
0.331
|
5.233
|
<0.001
|
|
WSR Z
|
1.705 (1.190 – 2.443)
|
0.313
|
2.905
|
0.004
|
|
novel [Held Out]
|
0.962 (0.872 – 1.062)
|
0.048
|
-0.769
|
0.442
|
|
outgroup [CSU LA]
|
0.917 (0.621 – 1.353)
|
0.182
|
-0.438
|
0.661
|
|
outgroup [UCLA]
|
0.444 (0.298 – 0.661)
|
0.090
|
-4.002
|
<0.001
|
|
desirability Z
|
1.197 (1.142 – 1.254)
|
0.029
|
7.514
|
<0.001
|
|
WSR Z * novel [Held Out]
|
0.960 (0.882 – 1.044)
|
0.041
|
-0.959
|
0.338
|
|
WSR Z * outgroup [CSU LA]
|
1.086 (0.646 – 1.824)
|
0.287
|
0.310
|
0.756
|
|
WSR Z * outgroup [UCLA]
|
0.834 (0.494 – 1.407)
|
0.223
|
-0.681
|
0.496
|
novel [Held Out] * outgroup [CSU LA]
|
0.986 (0.863 – 1.126)
|
0.067
|
-0.214
|
0.831
|
novel [Held Out] * outgroup [UCLA]
|
1.095 (0.955 – 1.256)
|
0.076
|
1.302
|
0.193
|
(WSR Z * novel [Held Out]) * outgroup [CSU LA]
|
0.985 (0.862 – 1.125)
|
0.067
|
-0.224
|
0.823
|
(WSR Z * novel [Held Out]) * outgroup [UCLA]
|
1.058 (0.926 – 1.210)
|
0.072
|
0.830
|
0.407
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.75
|
|
τ00 trait
|
0.34
|
|
τ11 subID.WSR.Z
|
1.56
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
τ11 trait.outgroupCSU LA
|
0.47
|
|
τ11 trait.outgroupUCLA
|
0.63
|
|
ρ01
|
0.11
|
|
|
0.59
|
|
|
-0.89
|
|
|
-0.16
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.119 / NA
|
ggpredict(m, c("WSR.Z", "novel", "outgroup")) %>% plot(show.title=F) + xlab("Similarity-Weighted Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="WSR.Z [all]"` to get smooth plots.

Cross-Validated Expectations
m <- glmer( ingChoiceN ~ predicted.Z * novel + desirability.Z + ( predicted.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.632 (1.416 – 1.880)
|
0.118
|
6.767
|
<0.001
|
|
predicted Z
|
1.573 (1.309 – 1.891)
|
0.148
|
4.822
|
<0.001
|
|
novel [Held Out]
|
0.988 (0.937 – 1.042)
|
0.027
|
-0.448
|
0.654
|
|
desirability Z
|
1.260 (1.182 – 1.344)
|
0.041
|
7.055
|
<0.001
|
predicted Z * novel [Held Out]
|
0.962 (0.912 – 1.016)
|
0.026
|
-1.397
|
0.162
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.53
|
|
τ00 trait
|
0.13
|
|
τ11 subID.predicted.Z
|
0.64
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.predicted.Z
|
0.25
|
|
ρ01 subID.novelHeld Out
|
-0.39
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.071 / NA
|
Dot-Product Expected Ratings
m <- glmer( ingChoiceN ~ er.Z * novel + desirability.Z + ( er.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.581 (1.322 – 1.891)
|
0.144
|
5.018
|
<0.001
|
|
er Z
|
1.409 (1.117 – 1.776)
|
0.167
|
2.895
|
0.004
|
|
novel [Held Out]
|
0.979 (0.928 – 1.032)
|
0.026
|
-0.791
|
0.429
|
|
desirability Z
|
1.264 (1.187 – 1.347)
|
0.041
|
7.229
|
<0.001
|
|
er Z * novel [Held Out]
|
0.962 (0.912 – 1.015)
|
0.026
|
-1.409
|
0.159
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.02
|
|
τ00 trait
|
0.13
|
|
τ11 subID.er.Z
|
1.95
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.er.Z
|
-0.01
|
|
ρ01 subID.novelHeld Out
|
-0.42
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.048 / NA
|
Linear Trend of Self-Descriptiveness
m <- glmer( ingChoiceN ~ slope.Z * novel + desirability.Z + ( slope.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.645 (1.373 – 1.970)
|
0.151
|
5.406
|
<0.001
|
|
slope Z
|
1.526 (1.194 – 1.950)
|
0.191
|
3.378
|
0.001
|
|
novel [Held Out]
|
0.980 (0.930 – 1.033)
|
0.026
|
-0.743
|
0.457
|
|
desirability Z
|
1.262 (1.184 – 1.345)
|
0.041
|
7.180
|
<0.001
|
slope Z * novel [Held Out]
|
0.960 (0.910 – 1.013)
|
0.026
|
-1.479
|
0.139
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.02
|
|
τ00 trait
|
0.13
|
|
τ11 subID.slope.Z
|
2.13
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.slope.Z
|
-0.06
|
|
ρ01 subID.novelHeld Out
|
-0.50
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.064 / NA
|
All Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveAllSE.Z * novel + desirability.Z + ( neighAveAllSE.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.701 (1.487 – 1.945)
|
0.116
|
7.756
|
<0.001
|
|
neighAveAllSE Z
|
1.241 (1.118 – 1.378)
|
0.066
|
4.052
|
<0.001
|
|
novel [Held Out]
|
0.988 (0.937 – 1.042)
|
0.027
|
-0.450
|
0.653
|
|
desirability Z
|
1.265 (1.187 – 1.347)
|
0.041
|
7.282
|
<0.001
|
neighAveAllSE Z * novel [Held Out]
|
0.983 (0.931 – 1.037)
|
0.027
|
-0.629
|
0.530
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.63
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveAllSE.Z
|
0.36
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.neighAveAllSE.Z
|
-0.08
|
|
ρ01 subID.novelHeld Out
|
-0.56
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29329
|
|
Marginal R2 / Conditional R2
|
0.031 / NA
|
Outwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveOutSE.Z * novel + desirability.Z + ( neighAveOutSE.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.724 (1.509 – 1.968)
|
0.117
|
8.038
|
<0.001
|
|
neighAveOutSE Z
|
1.204 (1.104 – 1.314)
|
0.053
|
4.190
|
<0.001
|
|
novel [Held Out]
|
0.981 (0.930 – 1.035)
|
0.027
|
-0.687
|
0.492
|
|
desirability Z
|
1.265 (1.188 – 1.347)
|
0.041
|
7.333
|
<0.001
|
neighAveOutSE Z * novel [Held Out]
|
0.980 (0.928 – 1.035)
|
0.027
|
-0.732
|
0.464
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.64
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveOutSE.Z
|
0.24
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.neighAveOutSE.Z
|
-0.07
|
|
ρ01 subID.novelHeld Out
|
-1.00
|
|
N subID
|
200
|
|
N trait
|
147
|
|
Observations
|
29060
|
|
Marginal R2 / Conditional R2
|
0.028 / NA
|
Inwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveInSE.Z * novel + desirability.Z + ( neighAveInSE.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.711 (1.495 – 1.959)
|
0.118
|
7.778
|
<0.001
|
|
neighAveInSE Z
|
1.168 (1.074 – 1.269)
|
0.050
|
3.647
|
<0.001
|
|
novel [Held Out]
|
0.984 (0.933 – 1.038)
|
0.027
|
-0.581
|
0.561
|
|
desirability Z
|
1.264 (1.186 – 1.348)
|
0.041
|
7.177
|
<0.001
|
neighAveInSE Z * novel [Held Out]
|
0.980 (0.928 – 1.034)
|
0.027
|
-0.749
|
0.454
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.67
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveInSE.Z
|
0.20
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.neighAveInSE.Z
|
0.05
|
|
ρ01 subID.novelHeld Out
|
-0.61
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29282
|
|
Marginal R2 / Conditional R2
|
0.024 / NA
|
Does generalization depend on outdegree?
Similarity-Weighted Self-Evaluations
m <- glmer( ingChoiceN ~ WSR.Z * novel * outDegree.Z + desirability.Z + ( WSR.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.674 (1.379 – 2.031)
|
0.165
|
5.213
|
<0.001
|
|
WSR Z
|
1.895 (1.452 – 2.472)
|
0.257
|
4.710
|
<0.001
|
|
novel [Held Out]
|
0.984 (0.933 – 1.038)
|
0.027
|
-0.595
|
0.552
|
|
outDegree Z
|
1.058 (0.985 – 1.137)
|
0.039
|
1.542
|
0.123
|
|
desirability Z
|
1.213 (1.133 – 1.299)
|
0.042
|
5.549
|
<0.001
|
|
WSR Z * novel [Held Out]
|
0.963 (0.912 – 1.018)
|
0.027
|
-1.326
|
0.185
|
|
WSR Z * outDegree Z
|
1.070 (1.033 – 1.109)
|
0.019
|
3.790
|
<0.001
|
novel [Held Out] * outDegree Z
|
1.025 (0.971 – 1.082)
|
0.028
|
0.896
|
0.370
|
(WSR Z * novel [Held Out]) * outDegree Z
|
0.962 (0.910 – 1.016)
|
0.027
|
-1.388
|
0.165
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.26
|
|
τ00 trait
|
0.12
|
|
τ11 subID.WSR.Z
|
2.58
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.WSR.Z
|
0.06
|
|
ρ01 subID.novelHeld Out
|
0.81
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.122 / NA
|
ggpredict(m, c("WSR.Z", "novel")) %>% plot(show.title=F) + xlab("Similarity-Weighted Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="WSR.Z [all]"` to get smooth plots.

Cross-Validated Expectations
m <- glmer( ingChoiceN ~ predicted.Z * novel * outDegree.Z + desirability.Z + ( predicted.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.652 (1.435 – 1.902)
|
0.119
|
6.975
|
<0.001
|
|
predicted Z
|
1.565 (1.303 – 1.881)
|
0.147
|
4.783
|
<0.001
|
|
novel [Held Out]
|
0.989 (0.938 – 1.043)
|
0.027
|
-0.401
|
0.689
|
|
outDegree Z
|
1.049 (0.973 – 1.131)
|
0.040
|
1.254
|
0.210
|
|
desirability Z
|
1.230 (1.145 – 1.321)
|
0.045
|
5.687
|
<0.001
|
predicted Z * novel [Held Out]
|
0.961 (0.911 – 1.014)
|
0.026
|
-1.450
|
0.147
|
|
predicted Z * outDegree Z
|
1.070 (1.034 – 1.108)
|
0.019
|
3.827
|
<0.001
|
novel [Held Out] * outDegree Z
|
1.019 (0.967 – 1.075)
|
0.028
|
0.709
|
0.478
|
(predicted Z * novel [Held Out]) * outDegree Z
|
0.961 (0.910 – 1.014)
|
0.027
|
-1.449
|
0.147
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.52
|
|
τ00 trait
|
0.13
|
|
τ11 subID.predicted.Z
|
0.64
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.predicted.Z
|
0.27
|
|
ρ01 subID.novelHeld Out
|
-0.44
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.072 / NA
|
Dot-Product Expected Ratings
m <- glmer( ingChoiceN ~ er.Z * novel * outDegree.Z + desirability.Z + ( er.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.598 (1.337 – 1.909)
|
0.145
|
5.150
|
<0.001
|
|
er Z
|
1.423 (1.129 – 1.794)
|
0.168
|
2.986
|
0.003
|
|
novel [Held Out]
|
0.980 (0.929 – 1.033)
|
0.027
|
-0.750
|
0.453
|
|
outDegree Z
|
1.062 (0.986 – 1.143)
|
0.040
|
1.591
|
0.112
|
|
desirability Z
|
1.227 (1.144 – 1.316)
|
0.044
|
5.695
|
<0.001
|
|
er Z * novel [Held Out]
|
0.960 (0.910 – 1.013)
|
0.026
|
-1.475
|
0.140
|
|
er Z * outDegree Z
|
1.071 (1.034 – 1.109)
|
0.019
|
3.852
|
<0.001
|
novel [Held Out] * outDegree Z
|
1.022 (0.968 – 1.078)
|
0.028
|
0.782
|
0.434
|
(er Z * novel [Held Out]) * outDegree Z
|
0.960 (0.908 – 1.014)
|
0.027
|
-1.465
|
0.143
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.02
|
|
τ00 trait
|
0.13
|
|
τ11 subID.er.Z
|
1.94
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.er.Z
|
-0.01
|
|
ρ01 subID.novelHeld Out
|
-0.40
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.052 / NA
|
Linear Trend of Self-Descriptiveness
m <- glmer( ingChoiceN ~ slope.Z * novel * outDegree.Z + desirability.Z + ( slope.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.664 (1.390 – 1.992)
|
0.153
|
5.553
|
<0.001
|
|
slope Z
|
1.540 (1.206 – 1.968)
|
0.192
|
3.458
|
0.001
|
|
novel [Held Out]
|
0.981 (0.930 – 1.035)
|
0.027
|
-0.703
|
0.482
|
|
outDegree Z
|
1.063 (0.987 – 1.144)
|
0.040
|
1.623
|
0.105
|
|
desirability Z
|
1.224 (1.141 – 1.313)
|
0.044
|
5.646
|
<0.001
|
slope Z * novel [Held Out]
|
0.958 (0.908 – 1.012)
|
0.026
|
-1.534
|
0.125
|
|
slope Z * outDegree Z
|
1.071 (1.034 – 1.109)
|
0.019
|
3.826
|
<0.001
|
novel [Held Out] * outDegree Z
|
1.020 (0.967 – 1.076)
|
0.028
|
0.722
|
0.470
|
(slope Z * novel [Held Out]) * outDegree Z
|
0.962 (0.910 – 1.016)
|
0.027
|
-1.379
|
0.168
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.01
|
|
τ00 trait
|
0.12
|
|
τ11 subID.slope.Z
|
2.12
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.slope.Z
|
-0.05
|
|
ρ01 subID.novelHeld Out
|
-0.49
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.068 / NA
|
All Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveAllSE.Z * novel * outDegree.Z + desirability.Z + ( neighAveAllSE.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.702 (1.490 – 1.945)
|
0.116
|
7.818
|
<0.001
|
|
neighAveAllSE Z
|
1.279 (1.151 – 1.420)
|
0.068
|
4.594
|
<0.001
|
|
novel [Held Out]
|
0.987 (0.936 – 1.041)
|
0.027
|
-0.475
|
0.635
|
|
outDegree Z
|
1.072 (0.997 – 1.153)
|
0.040
|
1.869
|
0.062
|
|
desirability Z
|
1.222 (1.140 – 1.309)
|
0.043
|
5.651
|
<0.001
|
neighAveAllSE Z * novel [Held Out]
|
0.975 (0.923 – 1.029)
|
0.027
|
-0.920
|
0.357
|
neighAveAllSE Z * outDegree Z
|
1.089 (1.050 – 1.130)
|
0.020
|
4.593
|
<0.001
|
novel [Held Out] * outDegree Z
|
1.013 (0.960 – 1.069)
|
0.028
|
0.470
|
0.638
|
(neighAveAllSE Z * novel [Held Out]) * outDegree Z
|
0.946 (0.894 – 1.001)
|
0.027
|
-1.922
|
0.055
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.63
|
|
τ00 trait
|
0.12
|
|
τ11 subID.neighAveAllSE.Z
|
0.36
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.neighAveAllSE.Z
|
-0.07
|
|
ρ01 subID.novelHeld Out
|
-0.57
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29329
|
|
Marginal R2 / Conditional R2
|
0.036 / NA
|
Outwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveOutSE.Z * novel * outDegree.Z + desirability.Z + ( neighAveOutSE.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.722 (1.510 – 1.964)
|
0.116
|
8.094
|
<0.001
|
|
neighAveOutSE Z
|
1.252 (1.146 – 1.368)
|
0.056
|
4.989
|
<0.001
|
|
novel [Held Out]
|
0.982 (0.931 – 1.036)
|
0.027
|
-0.662
|
0.508
|
|
outDegree Z
|
1.054 (0.979 – 1.134)
|
0.039
|
1.395
|
0.163
|
|
desirability Z
|
1.228 (1.146 – 1.315)
|
0.043
|
5.814
|
<0.001
|
neighAveOutSE Z * novel [Held Out]
|
0.970 (0.917 – 1.025)
|
0.028
|
-1.085
|
0.278
|
neighAveOutSE Z * outDegree Z
|
1.089 (1.048 – 1.132)
|
0.021
|
4.375
|
<0.001
|
novel [Held Out] * outDegree Z
|
1.024 (0.970 – 1.081)
|
0.028
|
0.860
|
0.390
|
(neighAveOutSE Z * novel [Held Out]) * outDegree Z
|
0.958 (0.904 – 1.016)
|
0.029
|
-1.419
|
0.156
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.63
|
|
τ00 trait
|
0.12
|
|
τ11 subID.neighAveOutSE.Z
|
0.24
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.neighAveOutSE.Z
|
-0.07
|
|
ρ01 subID.novelHeld Out
|
-1.00
|
|
N subID
|
200
|
|
N trait
|
147
|
|
Observations
|
29060
|
|
Marginal R2 / Conditional R2
|
0.033 / NA
|
Inwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveInSE.Z * novel * outDegree.Z + desirability.Z + ( neighAveInSE.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.710 (1.494 – 1.957)
|
0.118
|
7.806
|
<0.001
|
|
neighAveInSE Z
|
1.186 (1.091 – 1.289)
|
0.051
|
4.001
|
<0.001
|
|
novel [Held Out]
|
0.986 (0.935 – 1.040)
|
0.027
|
-0.517
|
0.605
|
|
outDegree Z
|
1.063 (0.987 – 1.144)
|
0.040
|
1.607
|
0.108
|
|
desirability Z
|
1.223 (1.140 – 1.313)
|
0.044
|
5.600
|
<0.001
|
neighAveInSE Z * novel [Held Out]
|
0.974 (0.922 – 1.028)
|
0.027
|
-0.963
|
0.336
|
neighAveInSE Z * outDegree Z
|
1.078 (1.039 – 1.118)
|
0.020
|
4.032
|
<0.001
|
novel [Held Out] * outDegree Z
|
1.021 (0.968 – 1.077)
|
0.028
|
0.761
|
0.447
|
(neighAveInSE Z * novel [Held Out]) * outDegree Z
|
0.957 (0.904 – 1.013)
|
0.028
|
-1.527
|
0.127
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.67
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveInSE.Z
|
0.20
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.neighAveInSE.Z
|
0.06
|
|
ρ01 subID.novelHeld Out
|
-0.60
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29282
|
|
Marginal R2 / Conditional R2
|
0.028 / NA
|
Does generalization depend on indegree?
Similarity-Weighted Self-Evaluations
m <- glmer( ingChoiceN ~ WSR.Z * novel * inDegree.Z + desirability.Z + ( WSR.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.663 (1.370 – 2.020)
|
0.165
|
5.138
|
<0.001
|
|
WSR Z
|
1.890 (1.448 – 2.466)
|
0.257
|
4.682
|
<0.001
|
|
novel [Held Out]
|
0.984 (0.933 – 1.038)
|
0.027
|
-0.606
|
0.545
|
|
inDegree Z
|
0.982 (0.919 – 1.049)
|
0.033
|
-0.535
|
0.592
|
|
desirability Z
|
1.249 (1.173 – 1.329)
|
0.040
|
6.938
|
<0.001
|
|
WSR Z * novel [Held Out]
|
0.963 (0.912 – 1.017)
|
0.027
|
-1.341
|
0.180
|
|
WSR Z * inDegree Z
|
1.050 (1.014 – 1.086)
|
0.018
|
2.775
|
0.006
|
novel [Held Out] * inDegree Z
|
1.048 (0.995 – 1.105)
|
0.028
|
1.760
|
0.078
|
(WSR Z * novel [Held Out]) * inDegree Z
|
0.957 (0.907 – 1.010)
|
0.026
|
-1.581
|
0.114
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.26
|
|
τ00 trait
|
0.12
|
|
τ11 subID.WSR.Z
|
2.59
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.WSR.Z
|
0.06
|
|
ρ01 subID.novelHeld Out
|
0.79
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.120 / NA
|
ggpredict(m, c("WSR.Z", "novel")) %>% plot(show.title=F) + xlab("Similarity-Weighted Self-Descriptiveness") + ylab("Likelihood of Ingroup Choice") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="WSR.Z [all]"` to get smooth plots.

Cross-Validated Expectations
m <- glmer( ingChoiceN ~ predicted.Z * novel * inDegree.Z + desirability.Z + ( predicted.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.641 (1.425 – 1.891)
|
0.119
|
6.858
|
<0.001
|
|
predicted Z
|
1.577 (1.311 – 1.896)
|
0.148
|
4.839
|
<0.001
|
|
novel [Held Out]
|
0.990 (0.938 – 1.044)
|
0.027
|
-0.387
|
0.699
|
|
inDegree Z
|
0.981 (0.916 – 1.051)
|
0.034
|
-0.544
|
0.587
|
|
desirability Z
|
1.261 (1.181 – 1.346)
|
0.042
|
6.954
|
<0.001
|
predicted Z * novel [Held Out]
|
0.961 (0.910 – 1.014)
|
0.026
|
-1.460
|
0.144
|
|
predicted Z * inDegree Z
|
1.045 (1.011 – 1.081)
|
0.018
|
2.567
|
0.010
|
novel [Held Out] * inDegree Z
|
1.042 (0.989 – 1.098)
|
0.028
|
1.552
|
0.121
|
(predicted Z * novel [Held Out]) * inDegree Z
|
0.960 (0.910 – 1.012)
|
0.026
|
-1.506
|
0.132
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.53
|
|
τ00 trait
|
0.13
|
|
τ11 subID.predicted.Z
|
0.64
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.predicted.Z
|
0.26
|
|
ρ01 subID.novelHeld Out
|
-0.40
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.072 / NA
|
Dot-Product Expected Ratings
m <- glmer( ingChoiceN ~ er.Z * novel * inDegree.Z + desirability.Z + ( er.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.590 (1.330 – 1.901)
|
0.145
|
5.085
|
<0.001
|
|
er Z
|
1.418 (1.124 – 1.788)
|
0.168
|
2.949
|
0.003
|
|
novel [Held Out]
|
0.980 (0.929 – 1.033)
|
0.027
|
-0.761
|
0.447
|
|
inDegree Z
|
0.988 (0.923 – 1.058)
|
0.034
|
-0.338
|
0.735
|
|
desirability Z
|
1.263 (1.184 – 1.347)
|
0.042
|
7.071
|
<0.001
|
|
er Z * novel [Held Out]
|
0.961 (0.910 – 1.014)
|
0.026
|
-1.469
|
0.142
|
|
er Z * inDegree Z
|
1.049 (1.014 – 1.086)
|
0.018
|
2.763
|
0.006
|
novel [Held Out] * inDegree Z
|
1.046 (0.992 – 1.102)
|
0.028
|
1.673
|
0.094
|
(er Z * novel [Held Out]) * inDegree Z
|
0.961 (0.910 – 1.014)
|
0.026
|
-1.463
|
0.144
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.02
|
|
τ00 trait
|
0.13
|
|
τ11 subID.er.Z
|
1.95
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.er.Z
|
-0.01
|
|
ρ01 subID.novelHeld Out
|
-0.40
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.050 / NA
|
Linear Trend of Self-Descriptiveness
m <- glmer( ingChoiceN ~ slope.Z * novel * inDegree.Z + desirability.Z + ( slope.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.655 (1.382 – 1.983)
|
0.152
|
5.479
|
<0.001
|
|
slope Z
|
1.535 (1.201 – 1.962)
|
0.192
|
3.426
|
0.001
|
|
novel [Held Out]
|
0.981 (0.930 – 1.034)
|
0.027
|
-0.713
|
0.476
|
|
inDegree Z
|
0.988 (0.924 – 1.058)
|
0.034
|
-0.335
|
0.738
|
|
desirability Z
|
1.260 (1.182 – 1.345)
|
0.042
|
7.024
|
<0.001
|
slope Z * novel [Held Out]
|
0.959 (0.908 – 1.012)
|
0.026
|
-1.534
|
0.125
|
|
slope Z * inDegree Z
|
1.050 (1.015 – 1.087)
|
0.018
|
2.817
|
0.005
|
novel [Held Out] * inDegree Z
|
1.045 (0.991 – 1.101)
|
0.028
|
1.623
|
0.104
|
(slope Z * novel [Held Out]) * inDegree Z
|
0.961 (0.910 – 1.014)
|
0.026
|
-1.464
|
0.143
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.01
|
|
τ00 trait
|
0.13
|
|
τ11 subID.slope.Z
|
2.13
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.slope.Z
|
-0.05
|
|
ρ01 subID.novelHeld Out
|
-0.50
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.066 / NA
|
All Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveAllSE.Z * novel * inDegree.Z + desirability.Z + ( neighAveAllSE.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.701 (1.488 – 1.945)
|
0.116
|
7.777
|
<0.001
|
|
neighAveAllSE Z
|
1.262 (1.137 – 1.402)
|
0.068
|
4.355
|
<0.001
|
|
novel [Held Out]
|
0.989 (0.938 – 1.044)
|
0.027
|
-0.392
|
0.695
|
|
inDegree Z
|
0.990 (0.925 – 1.059)
|
0.034
|
-0.300
|
0.764
|
|
desirability Z
|
1.262 (1.184 – 1.345)
|
0.041
|
7.120
|
<0.001
|
neighAveAllSE Z * novel [Held Out]
|
0.976 (0.924 – 1.030)
|
0.027
|
-0.886
|
0.376
|
neighAveAllSE Z * inDegree Z
|
1.063 (1.026 – 1.101)
|
0.019
|
3.379
|
0.001
|
novel [Held Out] * inDegree Z
|
1.043 (0.990 – 1.099)
|
0.028
|
1.577
|
0.115
|
(neighAveAllSE Z * novel [Held Out]) * inDegree Z
|
0.949 (0.898 – 1.003)
|
0.027
|
-1.857
|
0.063
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.63
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveAllSE.Z
|
0.36
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.neighAveAllSE.Z
|
-0.07
|
|
ρ01 subID.novelHeld Out
|
-0.52
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29329
|
|
Marginal R2 / Conditional R2
|
0.033 / NA
|
Outwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveOutSE.Z * novel * inDegree.Z + desirability.Z + ( neighAveOutSE.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.722 (1.508 – 1.966)
|
0.116
|
8.035
|
<0.001
|
|
neighAveOutSE Z
|
1.215 (1.114 – 1.326)
|
0.054
|
4.374
|
<0.001
|
|
novel [Held Out]
|
0.983 (0.932 – 1.037)
|
0.027
|
-0.628
|
0.530
|
|
inDegree Z
|
0.986 (0.922 – 1.055)
|
0.034
|
-0.402
|
0.688
|
|
desirability Z
|
1.263 (1.185 – 1.346)
|
0.041
|
7.174
|
<0.001
|
neighAveOutSE Z * novel [Held Out]
|
0.973 (0.922 – 1.028)
|
0.027
|
-0.973
|
0.330
|
neighAveOutSE Z * inDegree Z
|
1.053 (1.017 – 1.091)
|
0.019
|
2.887
|
0.004
|
novel [Held Out] * inDegree Z
|
1.049 (0.995 – 1.105)
|
0.028
|
1.760
|
0.078
|
(neighAveOutSE Z * novel [Held Out]) * inDegree Z
|
0.946 (0.896 – 1.000)
|
0.026
|
-1.971
|
0.049
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.64
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveOutSE.Z
|
0.24
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.neighAveOutSE.Z
|
-0.07
|
|
ρ01 subID.novelHeld Out
|
-1.00
|
|
N subID
|
200
|
|
N trait
|
147
|
|
Observations
|
29060
|
|
Marginal R2 / Conditional R2
|
0.029 / NA
|
Inwards Neighboring Self-Evaluations
m <- glmer( ingChoiceN ~ neighAveInSE.Z * novel * inDegree.Z + desirability.Z + ( neighAveInSE.Z + novel | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
## boundary (singular) fit: see help('isSingular')
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.710 (1.494 – 1.957)
|
0.118
|
7.800
|
<0.001
|
|
neighAveInSE Z
|
1.200 (1.103 – 1.306)
|
0.052
|
4.222
|
<0.001
|
|
novel [Held Out]
|
0.985 (0.934 – 1.039)
|
0.027
|
-0.563
|
0.574
|
|
inDegree Z
|
0.994 (0.928 – 1.064)
|
0.035
|
-0.187
|
0.852
|
|
desirability Z
|
1.259 (1.181 – 1.344)
|
0.042
|
6.985
|
<0.001
|
neighAveInSE Z * novel [Held Out]
|
0.972 (0.920 – 1.027)
|
0.027
|
-1.011
|
0.312
|
neighAveInSE Z * inDegree Z
|
1.063 (1.025 – 1.102)
|
0.019
|
3.324
|
0.001
|
novel [Held Out] * inDegree Z
|
1.046 (0.992 – 1.102)
|
0.028
|
1.673
|
0.094
|
(neighAveInSE Z * novel [Held Out]) * inDegree Z
|
0.968 (0.916 – 1.023)
|
0.027
|
-1.153
|
0.249
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
0.66
|
|
τ00 trait
|
0.13
|
|
τ11 subID.neighAveInSE.Z
|
0.20
|
|
τ11 subID.novelHeld Out
|
0.00
|
|
ρ01 subID.neighAveInSE.Z
|
0.05
|
|
ρ01 subID.novelHeld Out
|
-0.56
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29282
|
|
Marginal R2 / Conditional R2
|
0.027 / NA
|
Backwards solution: Can you predict self-evaluations from similarity
to ingroup and outgroup choices?
Main effect
m <- lmer( scale(selfResp) ~ scale(inGsim) + scale(outGsim) + ( scale(inGsim) + scale(outGsim) | subID) + (1 | trait), data = fullTrain)
## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
## Model failed to converge with max|grad| = 0.0027381 (tol = 0.002, component 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "B", string.se = "SE", string.stat = "t", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
B
|
SE
|
t
|
p
|
|
(Intercept)
|
-0.020 (-0.096 – 0.057)
|
0.039
|
-0.505
|
0.614
|
|
inGsim
|
0.163 (0.113 – 0.214)
|
0.026
|
6.352
|
<0.001
|
|
outGsim
|
-0.024 (-0.069 – 0.021)
|
0.023
|
-1.043
|
0.297
|
|
Random Effects
|
|
σ2
|
0.68
|
|
τ00 subID
|
0.17
|
|
τ00 trait
|
0.09
|
|
τ11 subID.scale(inGsim)
|
0.05
|
|
τ11 subID.scale(outGsim)
|
0.05
|
|
ρ01 subID.scale(inGsim)
|
0.05
|
|
ρ01 subID.scale(outGsim)
|
-0.03
|
|
ICC
|
0.31
|
|
N subID
|
197
|
|
N trait
|
148
|
|
Observations
|
17362
|
|
Marginal R2 / Conditional R2
|
0.022 / 0.323
|
Moderated by condition
m <- lmer( scale(selfResp) ~ scale(inGsim)*outgroup + scale(outGsim)*outgroup + ( scale(inGsim) + scale(outGsim) | subID) + ( outgroup | trait), data = fullTrain)
## boundary (singular) fit: see help('isSingular')
## Warning: Model failed to converge with 1 negative eigenvalue: -2.9e+01
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "B", string.se = "SE", string.stat = "t", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
B
|
SE
|
t
|
p
|
|
(Intercept)
|
0.001 (-0.113 – 0.115)
|
0.058
|
0.021
|
0.983
|
|
inGsim
|
0.174 (0.095 – 0.252)
|
0.040
|
4.312
|
<0.001
|
|
outgroup [Not UCR]
|
0.024 (-0.122 – 0.170)
|
0.074
|
0.317
|
0.751
|
|
outgroup [UCLA]
|
-0.086 (-0.231 – 0.060)
|
0.074
|
-1.155
|
0.248
|
|
outGsim
|
-0.007 (-0.080 – 0.066)
|
0.037
|
-0.186
|
0.852
|
inGsim * outgroup [Not UCR]
|
0.031 (-0.070 – 0.131)
|
0.051
|
0.602
|
0.547
|
|
inGsim * outgroup [UCLA]
|
-0.042 (-0.139 – 0.055)
|
0.049
|
-0.854
|
0.393
|
outgroup [Not UCR] * outGsim
|
-0.076 (-0.176 – 0.024)
|
0.051
|
-1.485
|
0.138
|
|
outgroup [UCLA] * outGsim
|
0.021 (-0.076 – 0.118)
|
0.050
|
0.420
|
0.675
|
|
Random Effects
|
|
σ2
|
0.68
|
|
τ00 subID
|
0.17
|
|
τ00 trait
|
0.09
|
|
τ11 subID.scale(inGsim)
|
0.05
|
|
τ11 subID.scale(outGsim)
|
0.05
|
|
τ11 trait.outgroupNot UCR
|
0.00
|
|
τ11 trait.outgroupUCLA
|
0.00
|
|
ρ01
|
0.03
|
|
|
-0.01
|
|
|
-1.00
|
|
|
-0.77
|
|
N subID
|
197
|
|
N trait
|
148
|
|
Observations
|
17362
|
|
Marginal R2 / Conditional R2
|
0.036 / NA
|
ggpredict(m, c("inGsim","outgroup")) %>% plot(show.title=F) + xlab("Similarity to Ingroup Choices") + ylab("Self-Evaluation") + jtools::theme_apa()

Moderated by Group Homophily
People who mentally segregrate the groups also self-evaluate more
similar to their later group choices.
# m <- lmer( scale(selfResp) ~ scale(inGsim) * scale(groupHomoph) + scale(outGsim) + ( scale(inGsim) + scale(outGsim) | subID) + (1 | trait), data = fullTrain)
# summary(m)
# tidy(m,conf.int=TRUE,effects="fixed")
# ggpredict(m, c("inGsim","groupHomoph")) %>% plot(show.title=F) + xlab("Similarity to Ingroup Choices") + ylab("Self-Evaluation") + jtools::theme_apa()
Individual differences moderation of self-anchoring
More self-projection for those who perceive group as more
entitative?
m <- glmer( ingChoiceN ~ WSR.Z * Ent.Z + ( WSR.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.627 (1.335 – 1.983)
|
0.164
|
4.824
|
<0.001
|
|
WSR Z
|
1.857 (1.418 – 2.432)
|
0.255
|
4.501
|
<0.001
|
|
Ent Z
|
1.133 (0.934 – 1.373)
|
0.111
|
1.269
|
0.204
|
|
WSR Z * Ent Z
|
1.117 (0.862 – 1.448)
|
0.148
|
0.839
|
0.402
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.28
|
|
τ00 trait
|
0.17
|
|
τ11 subID.WSR.Z
|
2.60
|
|
ρ01 subID
|
0.07
|
|
ICC
|
0.55
|
|
N subID
|
199
|
|
N trait
|
148
|
|
Observations
|
29184
|
|
Marginal R2 / Conditional R2
|
0.058 / 0.577
|
More self-projection for those who have been at UCR for shorter
time?
m <- glmer( ingChoiceN ~ WSR.Z * Years + ( WSR.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
2.180 (1.315 – 3.614)
|
0.562
|
3.023
|
0.003
|
|
WSR Z
|
1.726 (0.891 – 3.344)
|
0.582
|
1.618
|
0.106
|
|
Years
|
0.881 (0.717 – 1.082)
|
0.092
|
-1.212
|
0.225
|
|
WSR Z * Years
|
1.037 (0.797 – 1.349)
|
0.139
|
0.271
|
0.786
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.28
|
|
τ00 trait
|
0.17
|
|
τ11 subID.WSR.Z
|
2.58
|
|
ρ01 subID
|
0.07
|
|
ICC
|
0.55
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.051 / 0.573
|
More self-projection for higher self-esteem?
Higher self-esteem people self-project more
m <- glmer( ingChoiceN ~ WSR.Z * SE + ( WSR.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
0.787 (0.556 – 1.114)
|
0.140
|
-1.349
|
0.177
|
|
WSR Z
|
1.404 (1.054 – 1.871)
|
0.206
|
2.317
|
0.020
|
|
SE
|
19.366 (6.316 – 59.377)
|
11.070
|
5.184
|
<0.001
|
|
WSR Z * SE
|
2.529 (1.602 – 3.993)
|
0.589
|
3.983
|
<0.001
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.26
|
|
τ00 trait
|
0.15
|
|
τ11 subID.WSR.Z
|
2.57
|
|
ρ01 subID
|
0.07
|
|
ICC
|
0.55
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.049 / 0.570
|
ggpredict(m, c("SE","WSR.Z")) %>% plot(show.title=F) + xlab("Self-Esteem") + ylab("Self-Evaluation") + jtools::theme_apa()
## Data were 'prettified'. Consider using `terms="SE [all]"` to get smooth plots.

Do people higher in social identification self-project more?
No evidence
m <- glmer( ingChoiceN ~ WSR.Z * MGIS + ( WSR.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
0.676 (0.284 – 1.611)
|
0.299
|
-0.883
|
0.377
|
|
WSR Z
|
1.177 (0.393 – 3.527)
|
0.659
|
0.291
|
0.771
|
|
MGIS
|
1.208 (1.006 – 1.451)
|
0.113
|
2.026
|
0.043
|
|
WSR Z * MGIS
|
1.097 (0.869 – 1.384)
|
0.130
|
0.776
|
0.438
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.20
|
|
τ00 trait
|
0.17
|
|
τ11 subID.WSR.Z
|
2.52
|
|
ρ01 subID
|
0.05
|
|
ICC
|
0.54
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.064 / 0.571
|
Do people higher in need for cognition self-project more?
m <- glmer( ingChoiceN ~ WSR.Z * NFC + ( WSR.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
3.682 (1.335 – 10.157)
|
1.906
|
2.518
|
0.012
|
|
WSR Z
|
1.527 (0.395 – 5.910)
|
1.054
|
0.613
|
0.540
|
|
NFC
|
0.809 (0.624 – 1.048)
|
0.107
|
-1.604
|
0.109
|
|
WSR Z * NFC
|
1.060 (0.752 – 1.493)
|
0.185
|
0.331
|
0.741
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.28
|
|
τ00 trait
|
0.17
|
|
τ11 subID.WSR.Z
|
2.61
|
|
ρ01 subID
|
0.07
|
|
ICC
|
0.55
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.048 / 0.573
|
Do people higher in need to belong project more?
m <- glmer( ingChoiceN ~ WSR.Z * NTB + ( WSR.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
1.923 (0.698 – 5.303)
|
0.995
|
1.264
|
0.206
|
|
WSR Z
|
2.010 (0.519 – 7.787)
|
1.389
|
1.010
|
0.312
|
|
NTB
|
0.952 (0.704 – 1.287)
|
0.147
|
-0.320
|
0.749
|
|
WSR Z * NTB
|
0.978 (0.651 – 1.468)
|
0.203
|
-0.109
|
0.914
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.29
|
|
τ00 trait
|
0.17
|
|
τ11 subID.WSR.Z
|
2.59
|
|
ρ01 subID
|
0.06
|
|
ICC
|
0.55
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.051 / 0.574
|
Do people higher in self-prototypicality self-project more?
m <- glmer( ingChoiceN ~ WSR.Z * Proto + ( WSR.Z | subID) + (1 | trait), data = fullTest, family = binomial, control = glmerControl(optimizer = "bobyqa",
optCtrl = list(maxfun = 100000)),
nAGQ = 1)
tab_model(m, collapse.ci = T, show.stat=T, show.r2 = T, show.se = T, string.pred = c("Fixed Effects"), string.est = "OR", string.se = "SE", string.stat = "z", digits = 3, emph.p = F, dv.labels = "Ingroup Prediction")
|
|
Ingroup Prediction
|
|
Fixed Effects
|
OR
|
SE
|
z
|
p
|
|
(Intercept)
|
0.851 (0.434 – 1.670)
|
0.293
|
-0.469
|
0.639
|
|
WSR Z
|
1.094 (0.445 – 2.691)
|
0.502
|
0.195
|
0.845
|
|
Proto
|
1.127 (0.999 – 1.272)
|
0.069
|
1.949
|
0.051
|
|
WSR Z * Proto
|
1.099 (0.935 – 1.291)
|
0.091
|
1.141
|
0.254
|
|
Random Effects
|
|
σ2
|
3.29
|
|
τ00 subID
|
1.21
|
|
τ00 trait
|
0.17
|
|
τ11 subID.WSR.Z
|
2.54
|
|
ρ01 subID
|
0.04
|
|
ICC
|
0.54
|
|
N subID
|
200
|
|
N trait
|
148
|
|
Observations
|
29331
|
|
Marginal R2 / Conditional R2
|
0.062 / 0.572
|